Vivian Cook_ Li Wei - The Cambridge Handbook of Linguistic Multi-Competence-Cambridge University Press (2016)

The Cambridge Handbook of Linguistic Multi-competence
How are two or more languages learned and contained in the same mind or the
same community? This handbook presents an up-to-date view of the concept of
multi-competence, exploring the research questions it has generated and the
methods that have been used to investigate it. The book brings together
psychologists, sociolinguists, Second Language Acquisition (SLA) researchers,
and language teachers from across the world to look at how multi-competence
relates to their own areas of study. This comprehensive, state-of-the-art
exploration of multi-competence research and ideas offers a powerful critique
of the values and methods of classical SLA research, and an exciting preview of
the future implications of multi-competence for research and thinking about
language. It is an essential reference for all those concerned with language
learning, language use and language teaching.
V I V I A N C O O K is Emeritus Professor of Applied Linguistics at Newcastle University
and Visiting Professor at the University of York. He previously taught Applied

Linguistics at Essex University, and EFL and Linguistics in London.
L I W E I is Chair of Applied Linguistics and Director of the UCL Centre for Applied
Linguistics, at UCL Institute of Education, University College London. He has

previously worked at Birkbeck College, University of London, Newcastle
University, and Beijing Normal University.
Published online by Cambridge University Press

cambridge handbooks in language and linguistics
Genuinely broad in scope, each handbook in this series provides a complete

state-of-the-field overview of a major sub-discipline within language
study and research. Grouped into broad thematic areas, the chapters in
each volume encompass the most important issues and topics within
each subject, offering a coherent picture of the latest theories and
findings. Together, the volumes will build into an integrated overview of
the discipline in its entirety.
Published titles
The Cambridge Handbook of Phonology, edited by Paul de Lacy
The Cambridge Handbook of Linguistic Code-switching, edited by Barbara E. Bullock
and Almeida Jacqueline Toribio
The Cambridge Handbook of Child Language, Second Edition, edited by Edith L.
Bavin and Letitia Naigles
The Cambridge Handbook of Endangered Languages, edited by Peter K. Austin and
Julia Sallabank
The Cambridge Handbook of Sociolinguistics, edited by Rajend Mesthrie
The Cambridge Handbook of Pragmatics, edited by Keith Allan and Kasia M.
Jaszczolt
The Cambridge Handbook of Language Policy, edited by Bernard Spolsky
The Cambridge Handbook of Second Language Acquisition, edited by Julia
Herschensohn and Martha Young-Scholten
The Cambridge Handbook of Biolinguistics, edited by Cedric Boeckx and Kleanthes
K. Grohmann
The Cambridge Handbook of Generative Syntax, edited by Marcel den Dikken
The Cambridge Handbook of Communication Disorders, edited by Louise
Cummings
The Cambridge Handbook of Stylistics, edited by Peter Stockwell and Sara
Whiteley
The Cambridge Handbook of Linguistic Anthropology, edited by N.J. Enfield, Paul
Kockelman and Jack Sidnell
The Cambridge Handbook of English Corpus Linguistics, edited by Douglas Biber and
Randi Reppen
The Cambridge Handbook of Bilingual Processing, edited by John W. Schwieter
The Cambridge Handbook of Learner Corpus Research, edited by Sylviane Granger,
Gaëtanelle Gilquin and Fanny Meunier
The Cambridge Handbook of English Historical Linguistics, edited by Merja Kytö and
Päivi Pahta
The Cambridge Handbook of Linguistic Multicompetence, edited by Li Wei and Vivian
Cook
Forthcoming
The Cambridge Handbook of Formal Semantics, edited by Maria Aloni and Paul
Dekker
The Cambridge Handbook of Linguistic Typology, edited by Alexandra Aikhenvald
and R. M. W. Dixon
The Cambridge Handbook of Morphology, edited by Andrew Hippisley and Greg
Stump

The Cambridge
Handbook of Linguistic
Multi-competence
Edited by
Vivian Cook and Li Wei

University Printing House, Cambridge CB2 8BS, United Kingdom
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107059214
© Cambridge University Press 2016
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2016
Printed in the United Kingdom by Clays, St Ives plc
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Cook, Vivian, 1940– editor. | Li, Wei, 1961 August 11– editor.
The Cambridge handbook of linguistic multi-competence / edited by
Vivian Cook and Li Wei.
Handbook of linguistic multi-competence
Cambridge ; New York: Cambridge University Press, [2016] |
Series: Cambridge Handbooks in Language and Linguistics
LCCN 2015041119 | ISBN 9781107059214 (hardback)
LCSH: Multilingualism – Handbooks, manuals, etc. | Communicative
competence – Handbooks, manuals, etc. | Second language acquisition –
Handbooks, manuals, etc. | Language awareness – Handbooks, manuals, etc. |
BISAC: LANGUAGE ARTS & DISCIPLINES / Linguistics / General.
LCC P115.4 .C36 2016 | DDC 404/.2–dc23
LC record available at http://lccn.loc.gov/2015041119
ISBN 978-1-107-05921-4 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.





Published titles
Julia Sallabank
Jaszczolt
K. Grohmann
Cummings
Whiteley
Randi Reppen
Päivi Pahta
Cook
Forthcoming
Dekker
and R. M. W. Dixon
Stump

The Cambridge
Multi-competence
Edited by


www.cambridge.org
LCC P115.4 .C36 2016 | DDC 404/.2–dc23
ISBN 978-1-107-05921-4 Hardback





Published titles
Julia Sallabank
Jaszczolt
K. Grohmann
Cummings
Whiteley
Randi Reppen
Päivi Pahta
Cook
Forthcoming
Dekker
and R. M. W. Dixon
Stump

The Cambridge
Multi-competence
Edited by


www.cambridge.org
LCC P115.4 .C36 2016 | DDC 404/.2–dc23
ISBN 978-1-107-05921-4 Hardback





Published titles
Julia Sallabank
Jaszczolt
K. Grohmann
Cummings
Whiteley
Randi Reppen
Päivi Pahta
Cook
Forthcoming
Dekker
and R. M. W. Dixon
Stump

The Cambridge
Multi-competence
Edited by


www.cambridge.org
LCC P115.4 .C36 2016 | DDC 404/.2–dc23
ISBN 978-1-107-05921-4 Hardback

Contents
List of figures page vii

List of tables viii
List of contributors x
Acknowledgments xii
1 Premises of multi-competence Vivian Cook 1

2 Research questions and methodology of multi-competence
Goro Murahata, Yoshiko Murahata and Vivian Cook 26
3 Multi-competence in second language acquisition: inroads
into the mainstream? Lourdes Ortega 50
4 Not through a glass darkly: refocusing the psycholinguistic
study of bilingualism through a “bivocal” lens Jyotsna Vaid
and Renata Meuter 77
5 Multilingualism research Rita Franceschini 97
6 Multi-competence and dynamic/complex systems Kees de Bot 125
7 Multi-competence and Dominant Language Constellation
Larissa Aronin 142
8 Consequences of multi-competence for sociolinguistic
research Li Wei 164
9 A usage-based account of multi-competence Joan Kelly Hall 183
10 Multi-competence and syntax Éva Berkes and Suzanne Flynn 206
11 Syntactic processing Leah Roberts 227
12 Language and cognition in bilinguals Annette M. B. de Groot 248
13 Gestures in multi-competence Amanda Brown 276
14 Pragmatic transfer in foreign language learners: a
multi-competence perspective I-Ru Su 298
15 Multi-competence and endangered language
revitalization Tracy Hirata-Edds and Lizette Peter 321

vi Contents
16 Multi-competence and first language attrition Bregtje Seton

and Monika S. Schmid 338
17 Cognitive consequences of multi-competence Panos
Athanasopoulos 355
18 Space, motion and thinking for language Anna Ewert 376
19 Multi-competence and personality Jean-Marc Dewaele 403
20 Multi-competence as a creative act: ramifications of the
multi-competence paradigm for creativity research and
creativity-fostering education Anatoliy V. Kharkhurin 420
21 Multi-competence and language teaching Virginia M. Scott 445
22 Multi-competence and emotion Jean-Marc Dewaele 461
23 Multi-competence and English as
a lingua franca Ian MacKenzie 478
24 A critical reaction from second language acquisition
research David Singleton 502
25 Questions of multi-competence: a written interview
Guillaume Thierry 521
26 Epilogue: multi-competence and the Translanguaging
Instinct Li Wei 533
Bibliography of ‘multi-competence’ (and related topics) Goro Murahata,

Yoshiko Murahata and Vivian Cook 544
Index 559

Figures
1.1 The integration continuum of possible relationships

in multi-competence page 9
5.1 The mutual approximation of the two terms
multi-competence and multilingualism 110
7.1 DLC in Israel: Rose, Arabic sector school 147
7.2 DLC in Israel: Iman, Arabic sector school 148
7.3 DLC in Israel: Liti, Jewish sector school 149
7.4 DLC in Israel: user of Russian, Hebrew and English 150
12.1 The Revised Hierarchical Model 257
12.2 The Shared Distributed Asymmetrical Model 259
13.1 A native Japanese speaker describing a scene from the
cartoon Canary Row 279
14.1 English and Chinese native speakers’ use of different
types of request strategies 305
14.2 The frequencies of different types of request strategies
among ENSs and EFL learners 306
14.3 The frequencies of different types of request strategies
among CNSs and EFL learners 308
14.4 Mean number of apology strategies across status
conditions in the L1 control groups 312
14.5 Mean number of apology strategies across status
conditions in CNSs and EFL learners 315

Tables
3.1 Study designs for the investigation of cross-linguistic

influence in SLA page 53
4.1 Summary of journal outlets in psychology of language
citing Cook (1991) and Cook (1992) 83
4.2 Corpus analysis for the impact of Cook (1991, 1992)
and Grosjean (1985, 1989) using ProQuest 84
6.1 Six meanings of ‘language’ 129
7.1 Inter-related perspectives and concepts 146
7.2 The connection between multi-competence and DLC
and their role in research 155
7.3 Features for investigation of multi-competence
manifested in a DLC, with examples 158
10.1 CP-related features in background language, from
selected studies 215
14.1 Examples of supportive moves identified in the study 304
14.2 Mean number of conventionally indirect strategies in
relation to social variables in the L1 control groups 305
14.3 English requests by ENSs and Chinese EFL learners 306
relation to social variables in the English DCT 307
14.5 Chinese requests by CNSs and Chinese EFL learners 308
relation to social variables in the Chinese DCT 309
14.7 Intermediate EFL learners’ requests in English
and Chinese 309
14.8 Advanced EFL learners’ requests in English and Chinese 309
14.9 Distribution of apology strategies in the L1 control
groups 311
14.10 Mean number of apology strategies in relation to social
variables in the L1 control groups 312
variables in the English DCT 313

List of tables ix

variables in the Chinese DCT 315
14.13 Intermediate EFL learners’ mean number of strategies
in relation to social variables in English and Chinese
apologies 316
14.14 Advanced EFL learners’ mean number of strategies
in relation to social variables in English and Chinese
apologies 317

Contributors
Larissa Aronin Associate Professor, Oranim Academic College of

Education, Israel; Visiting Research Fellow, Trinity College Dublin,
Ireland
Panos Athanasopoulos Professor, Linguistics and English Language,
Lancaster University, England
Éva Berkes University of Applied Sciences of Burgenland, Austria
Amanda Brown Associate Professor, Languages, Literatures, and
Linguistics, University of Syracuse, USA
Vivian Cook Emeritus Professor of Applied Linguistics, Newcastle
University, England
Kees de Bot Professor, Department of Applied Linguistics, University of
Groningen, The Netherlands; Professor of Applied Linguistics,
University of Pannonia, Hungary
Annette M. B. de Groot Professor of Experimental Psycholinguistics,
University of Amsterdam, The Netherlands
Jean-Marc Dewaele Professor in Applied Linguistics and Multilingualism,
Birkbeck, University of London
Anna Ewert Assistant Professor, Faculty of English, Adam Mickiewicz
University, Poland
Suzanne Flynn Professor of Foreign Languages and Linguistics,
Department of Linguistics and Philosophy, Massachusetts Institute of
Technology, USA
Rita Franceschini Professor, Competence Centre for Language Studies,
University of Bolzano/Bozen, Italy
Joan Kelly Hall Professor of Applied Linguistics, Department of Applied
Linguistics, Pennsylvania State University, USA
Tracy Hirata-Edds Lecturer, Applied English Center, University of
Kansas, USA

List of contributors xi
Anatoliy V. Kharkhurin Associate Professor of Psychology, Department

of International Studies, American University of Sharjah, United Arab
Emirates
Li Wei Professor of Applied Linguistics, UCL Institute of Education,
University College London
Ian MacKenzie Maı̂tre d’enseignement et de recherche, Faculté de
traduction et d’interprétation, Geneva University, Switzerland
Renata Meuter Associate Professor, School of Psychology and
Counselling, Queensland University of Technology, Australia
Goro Murahata Professor of Applied English Linguistics, University of
Miyazaki, Japan
Yoshiko Murahata Instructor of English, University of Miyazaki, Japan
Lourdes Ortega Professor, Department of Linguistics, Georgetown
University, USA
Lizette Peter Associate Professor, Department of Curriculum and
Teaching, University of Kansas, USA
Leah Roberts Professor, Department of Education, University of York,
England
Monika S. Schmid Professor, Language and Linguistics, University of
Essex, England, and University of Groningen, The Netherlands
Virginia M. Scott Professor of French and Applied Linguistics, Vanderbilt
University, USA
Bregtje Seton Centre for Language and Cognition, University of
Groningen, The Netherlands
David Singleton Fellow Emeritus, Trinity College Dublin, Ireland;
Professor of Applied Linguistics, University of Pannonia, Hungary
I-Ru Su Associate Professor, Department of Foreign Languages and
Literature, National Tsing Hua University, Taiwan
Guillaume Thierry Professor of Cognitive Neuroscience, Ysgol Seicoleg |
School of Psychology, Prifysgol Bangor | Bangor University, Wales
Jyotsna Vaid Professor of Cognition and Cognitive Neuroscience,
Department of Psychology, Texas A&M University, USA

Acknowledgments
We would like to thank our amazing contributors, who have seen far more
scope in the idea of multi-competence than Vivian Cook ever dreamt of
when he proposed it at AILA in Thessaloniki in 1989 as a contribution to
a symposium about language input. They have shown us some of the
wealth of possibilities that this perspective provides for language-related
fields. Vivian Cook would also like to express his appreciation of the
generation of former students who have contributed so much to its devel-
opment: Fatimah Almutrafi, Panos Athanasopoulos, Benedetta Bassetti,
Mohammad Ibrahim, Chise Kasai, Yoshiko Murahata, Miho Sasaki, Petra
Schoofs, Jun Takahashi, Yuki Tokumaru and many others. Li Wei would
like to thank Zhu Hua, Jean-Marc Dewaele, David Green, Ofelia Garcia,
amongst others, who have discussed the ideas of multi-competence and
translanguaging with him.
As always for Vivian Cook, the book would never have seen the light of
day without the musical accompaniment of Ibrahim Maalouf, Marcin
Wasilewski and Seb Rochford. And for Li Wei, the multi-competence of
his young sons never ceases to amaze him.

1
Premises of
multi-competence
Vivian Cook
This chapter introduces the concept of linguistic multi-competence

and sets the scene for the rest of the book. It looks first at issues of
definition and then at three premises that have become part and
parcel of multi-competence. The aim is to examine the ideas under-
lying multi-competence rather than to present new views of multi-
competence or to summarise existing research, to be tackled in
Chapter 2.
1.1 Monolingual and bilingual perspective
There are two alternative ways of looking at people who speak more
than one language. On the one hand there is the monolingual perspective
that sees second language (L2) users from the point of view of the
monolingual first language (L1) user. In this case the second language
is added on to the speaker’s first language, something extra; the L2
user’s proficiency in the second language is measured against the sole
language of the monolingual; ideally the L2 user would speak the
second language just like a native speaker. The research questions and
methodology in classical second language acquisition (SLA) research are
mostly concerned with this monolingual perspective and try to account
for L2 users’ lack of success in learning how to speak like a monolingual
L1 user.
On the other hand there is the bilingual perspective that sees L2 users from
the point of view of the person who speaks two or more languages. From
this angle, the other languages are part of the L2 user’s total language
system, each language potentially differing from that of someone who
speaks it as a monolingual. It is beside the point whether the L2 user’s
final ability is identical to that of a monolingual native speaker.
Bilingualism and multilingualism research have mostly asked questions
https://doi.org/10.1017/CBO9781107425965.001 Published online by Cambridge University Press

2 VIVIAN COOK
about how L2 users use the other languages and how the languages con-
nect in multilingual communities, not about how L2 users compare with
monolingual individuals and communities.
One interpretation of the bilingual perspective is captured by the
notion of multi-competence, glossed here as ‘the overall system of a
mind or a community that uses more than one language’. Multi-
competence thus covers the knowledge and use of two or more lan-
guages by the same individual or the same community. At some level,
all the languages form part of one overall system, with complex and
shifting relationships between them, affecting the first language as
well as the others.
1.2 Defining multi-competence
Let us start with the conceptual history of multi-competence. Franceschini

(2011) interprets the history with a slightly different focus, largely as a
development from a psychological generative tradition to a dynamic socio-
linguistics of multilingualism.
(i) ‘the compound state of a mind with two grammars’

The term ‘multi-competence’ was first used in an SLA context to mean
‘the compound state of a mind with two grammars’ (Cook 1991), partly
to complement the term ‘interlanguage’, which refers solely to the L2
component in the bilingual mind, ignoring the L1 component. Multi-
competence saw second language acquisition as involving the whole
mind of the L2 user, not just the second language; multi-competence
included all language-related aspects of the mind. The word grammar in
the original definition was intended in the Chomskyan sense of knowl-
edge of language – ‘we call the theory of the state attained its [the language
faculty’s] grammar’ (Chomsky and Lasnik 1993). That is to say, grammar
includes all aspects of linguistic knowledge such as vocabulary and pho-
nology, not just syntax alone. However, this Chomskyan sense of
grammar turned out to be misleading as it led some researchers into
thinking that multi-competence was only about syntax rather than the
totality of language knowledge.
(ii) ‘the knowledge of more than one language in the same mind
or the same community’
The definition of multi-competence was later modified to ‘the knowledge
of more than one language in the same mind’ (Cook 2003) and, more
recently, to ‘the knowledge of more than one language in the same mind
or the same community’ (Cook 2012). These changes affected the original
definition by:

Premises of multi-competence 3
– making clear that multi-competence is not confined to syntax but

includes the lexicon, phonology etc.
– going beyond the second language to other languages the L2 user may
know – the relationship between three or more languages is equally a
matter of multi-competence
– extending the concept beyond the psychological construct of the
mind of the individual to the sociological construct of the ‘multi-
competence of the community’ (Brutt-Griffler 2002), treating the
diverse languages of the community as a coherent whole rather than
separately.
This 2012 definition (ii) is the one that most of the contributors in
this volume refer to, apart from Hall (Chapter 9) who uses the 1991 defini-
tion (i).
(iii) ‘the overall system of a mind or a community that uses more

than one language’
Yet the change from ‘the compound state’ in definition (i) to ‘the knowledge
of more than one language’ in definition (ii) has the unintended conse-
quence of implying a static view of language as knowledge rather than a
social definition of language or a multifaceted view of language
and language use. On reflection, a preferable working definition is ‘the
overall system of a mind or a community that uses more than one lan-
guage’. This changes ‘knowledge’ to the more neutral ‘system’, does not
confine multi-competence to language alone, brings in language use and
implies that language is not separate from the rest of the mind. This defini-
tion is not fully acceptable to all the contributors to this volume and it still
leaves the concepts of ‘system’ and ‘community’ open to interpretation.
As the multi-competence approach developed and broadened, it became
evident that it was more a perspective from which to view the acquisition
and use of multiple languages than a theory or a model. Multi-competence
is a way of looking at things from another angle rather than of exploring
the implications and contradictions within the same perspective, ‘revolu-
tionary’ rather than ‘normal’ science (Kuhn 1962). The monolingual per-
spective yields SLA research questions and methods that are inextricably
linked to monolingual native speakers; the multi-competence perspective
relates its questions and methods to L2 users. Thus many classic ‘normal’
issues are neither here nor there for multi-competence research. The
failure of L2 users to speak like natives, the inability of L2 users who
start learning at an older age to sound like natives, the L2 user’s lack of
elements of Universal Grammar possessed by natives – none of these are
meaningful from a multi-competence perspective. The monolingual
perspective in essence restricts the field of SLA research to enumerating
the similarities and dissimilarities between L2 users and native speakers.
If L2 users are independent persons in their own right rather than

4 VIVIAN COOK
the shadows of native speakers, the comparison between L2 users and

monolingual native speakers is about as revealing as, say, discussing how
apples resemble pears, of little interest for those concerned with the
distinctive qualities of apples.
1.3 The second language user
The other term that seemed to go naturally with multi-competence was

L2 user, meaning ‘people who know and use a second language at any level’
(Cook 2012), rather than L2 learner or bilingual. It seemed better to treat
people as users of a language whatever their level rather than as learners
who would never be complete: ‘SLA researchers often portray develop-
ment as a transitional state that is (or should be) ever changing towards the
target’ (Ortega 2009, p. 140). It would be insulting to call Björn Ulvaeus of
Abba, Joseph Conrad the novelist or Aung San Suu Kyi the politician L2
learners of English when they are capable of using their second language
to function in their respective ways at a level beyond the dreams of most
monolingual native speakers. Calling people L2 learners confirms their
subordinate status as learners for the rest of their days.
The term L2 learners can then be reserved for people who are learning
another language but are not using it, in other words those whose sole
purpose is learning the language, say Chinese children learning English in
Shanghai. Of course some L2 learners go on to become L2 users in later life
and some L2 learners use the language for real-world functions in the
classroom when they step outside, like Chinese students in Newcastle
upon Tyne. In other words a particular individual may be an L2 learner
or an L2 user at different times in their life or indeed at different times of
day. Classroom L2 learners at best are deferred L2 users. In practice this
virtually restricts L2 learner to students or pupils since people acquiring
another language outside education will almost always be using the sec-
ond language. And this necessarily raises the issue, not to be developed
here, of whether there are in effect two branches of SLA research, one
concerned with ‘natural’ acquisition and use, the other with teacher-
induced learning in classrooms, and so results from one branch do not
necessarily apply to the other.
In some ways the distinction between user and learner overlaps with
the traditional, slightly confusing, distinction in language teaching
between second language learners using a language for everyday living in
a country where it is the main community language, and foreign language
learners who are not learning a language for immediate use in a country
where it is spoken (Klein 1986). One problem with this distinction is
the conflation of function and location (Cook 2010): students at English-
medium universities in Saudi Arabia or the Netherlands may be using
English as a second language; overseas students at UK language schools

may be learning English only to have a qualification to show back home;

waiters use Spanish as a lingua franca in London (Block 2007), giving it a
second language function in an English-speaking country.
There were also reasons for minimising the use of the term bilingual.
Most people tend to assume that bilingual conveys Bloomfield’s maximal
meaning of ‘native-like control of two languages’ (Bloomfield 1933), rather
than Haugen’s minimal meaning of ‘the point where a speaker can first
produce complete meaningful utterances in the other language’ (Haugen
1953). Multi-competence did not assume that L2 users were at a high level
in the second language, particularly when, like Bloomfield, this is defined
in terms of likeness to native speakers. Multi-competence concerns
the mind of any user of a second language at any level of achievement.
There may well be a maximal level for the L2 user, sometimes called ‘the
successful L2 user’. But until a norm is set for L2 use that does not refer
to the native speaker, we don’t know what this might be. And it might
indeed vary considerably between the different users and uses of the
second language. Additionally bilingual conveys the notion that two lan-
guages are involved when there may be an indefinite number. The term
multilingual is perhaps closest in connotation to L2 user, not excluding more
than two languages and not hinting that language proficiency has to be
high.
Moving from the monolingual perspective that a human being knows
one language to the multi-competence perspective that all human beings
potentially, and some actually, know more than one language changes
the view of the whole landscape. Thus SLA theories as diverse as generative
grammar and usage-based acquisition can be conceived from both mono-
lingual and bilingual perspectives, as we see in the rest of this book. On
the one hand, say, ideas of innateness are seen from the angle that it is
normal to know more than one language (Cook 2009a). On the other hand,
usage needs to start from the total language input, not just that in one
language but in both languages, hinting that knowing and using only one
language is a form of deprivation, not so much linguistic deprivation
which fails to provide a child with crucial aspects of one language but
language deprivation which deprives them of a whole second language.
Multi-competence is not confined to psychological theories of the mind
but applies also to the networks of connectionist models, to generative
theories of language knowledge and to sociological models of social
interaction and practice: it is the perspective from which the languages
in the mind are viewed that matters, regardless of the theory involved.
Multi-competence alters the way in which people view the acquisition
and use of multiple languages, rather like the shift from seeing Short
Term Memory as boxes to seeing it as depth of processing (Craik and
Lockhart 1972), which essentially restated how the very same facts about
human memory were viewed. In part it leads to research with specifically
multi-competence aims, in part to reinterpreting existing research that

6 VIVIAN COOK
can be compatible with multi-competence, in part to a critique of SLA

research that is uninterpretable from a multi-competence perspective.
A logical problem arising out of this, addressed in Cook (2010), is the
meaning of second language and L2 (and indeed of language, to be discussed
below). Second language and L2 are not equivalent in meaning despite
most researchers’ habit of reading L2 aloud as second language. The word
second is ordinal counting for sequence – King Edward II came after King
Edward I – or for quality – a second-class degree is less valued than a first-class
degree. The number 2 on the other hand is cardinal counting of quantity –
two drinks, two Houses of Parliament. The meaning of L2/second in SLA
research could be any of these; Hammerberg (2010, p. 93) describes ‘the
linear model’ of counting languages involving ordinal counting in which
a second language comes chronologically after a first and a third language
after that: ‘Joseph Conrad’s first language was Polish, second language
French, third language English.’
Undoubtedly some of the overtones of the ‘quality’ ordinal meaning
carry across to SLA research; second is by and large not a good thing to
be – second-hand, second-rate, second-in-command.
Although the now discredited notions such as native speaker or mother
tongue speaker require us to identify ourselves according to our parental
language or language of infancy, even the alternatives such as L1 and L2
force us to identify a single language as receiving primacy in terms of
our time of acquisition or level of competence. (Canagarajah 2007, p. 16)
The letters of the alphabet are used in a similar ordinal fashion for defining
priority in putting airplane passengers into boarding Groups A, B or C or
marking essays as A, B . . . F, and so on.
In cardinal counting the meaning is more neutral: how many languages
you know – ‘Joseph Conrad spoke three languages, Polish, French and
English’ – rather than the order or priority between them, in linguistic
terms a synchronic state rather than a diachronic process. Similarly, think-
ing ‘cardinally’, the alphabet has 26 equal letters, that happen to occur in
an arbitrary order. Hammerberg’s (2010) alternative terminology of pri-
mary, secondary and tertiary languages still carries the ordinal overtones
of primary being superior and essential. Dewaele (Chapter 19, this volume)
uses the more neutral term LX to refer to any language beyond the first.
The academic discussion of first and second languages is also muddied
by the different ways in which countries define their first languages. In
Singapore schools for instance the first language is English; Chinese,
Bahasa Malaysia and Tamil, the mother tongues of most inhabitants, are
regarded as second languages. Another problem is where counting stops –
L3, L4, Ln. Most SLA books claim second subsumes later languages, whether
‘second (third, etc) languages and dialects’ (Doughty and Long 2003, p. 3)
or ‘the third or fourth language’ (Lightbown and Spada 2006, p. 204).
This simplification assumes that multi-competence with three or

more languages is just a more complex version of that with two languages
rather than something qualitatively different, strongly denied by those
interested in trilingualism and multilingualism who stress ‘the unique
properties that differentiate L2 from L3/Ln’ (Cabrelli Amaro, Flynn and
Rothman 2012, p. 3).
1.4 Three premises of multi-competence
A way of drawing out the implications of the multi-competence position

for second language acquisition is to derive three premises from the
developing stream of multi-competence-related work. They can be seen
as threads running through the following chapters, which the contribu-
tors are free to accept or reject.
The historical development of the concept of multi-competence has
perhaps been more a matter of teasing out and clarifying the implications
of the original proposal rather than of changing direction. The rest of
this chapter deals with three premises that seem to underlie multi-
competence. It was only at the beginning of the twenty-first century that
people began to use multi-competence to explore the research questions
to be discussed in Chapter 2, as it fitted in with the zeitgeist about the role
of the native speaker and with developing ideas about multilingualism.
Premise 1 Multi-competence concerns the total system for all

languages (L1, L2, Ln) in a single mind or community
and their inter-relationships
Despite the many books on bilingualism whose covers feature two heads
(Skutnabb-Kangas 1981; Romaine 1994; Pavlenko 2005; among others),
bilinguals do only have one.1 At the highest level of all, the languages
must be an inter-connected whole within a single mind, an eco-system of
mutual interdependence. At the same general level, a multi-lingual com-
munity is an interconnected network of different languages: in London in
2011, 6.5 per cent of the population spoke Polish, Panjabi, Urdu, Bengali
and Gujarati (Office for National Statistics 2012), not to mention the other
300 odd languages in the community (Baker and Eversley 2000); in
Vancouver in 2011, 57.7 per cent of the population spoke an immigrant
language at home (Statistics Canada 2012). The question is not how lin-
guistic enclaves function in isolation from each other but how the whole
city functions through multiple languages. To take an example of street
signs, it is not which language is used in which signs that matters so much
as how the street signs make up a total multi-competent system
(Cook 2013).
The description of L2 users and communities has thus in principle to
account for all the languages they use, both their first language and any
others, as part of one complex system. Isolating L2 syntax from L1 syntax

8 VIVIAN COOK
in the L2 user’s mind is a simplification for convenience of research. The

reality is the overall system that unites the first language and the other
language or languages of multi-competence. SLA research that ignores
the first language element is blind to the one inescapable feature of the
L2 user’s mind that distinguishes it from that of a monolingual – the first
language system. It is yin without yang. As Stern (1992, p. 282) puts it,
‘whether we like it or not, the new language is learnt on the basis of
a previous language’. Unless the presence of the first language is
acknowledged, second language acquisition research inevitably becomes
a footnote to first language acquisition.
It is an empirical question how and at what level the languages of
multi-competence separate in the mind or indeed whether it is mean-
ingful to attempt to separate them at all, as de Bot suggests in Chapter 6.
Cook (2009a) argues for an overall unified grammar in the mind as the
basis of Universal Grammar theory. The reverse question is whether
languages can be kept separate in the mind: can one be turned off
while the other is being used? Lambert posed the question in terms of
gating:
How is it that the bilingual is able to ‘gate out’ or set aside a whole
integrated linguistic system while functioning with a second one and a
moment later, if the situation calls for it, switch the process, activating the
previous inactive system and setting aside the previous active one?
(Lambert 1990, pp. 203–204)
An alternative is that, rather than one language being activated, the other
language is turned off, as in the Inhibitory Control Model (Green 1998),
leading to the emphasis on executive control in contemporary bilingual-
ism research (Bialystok 2009).
Turning to some evidence, if L2 users are shown pictures of objects
named in one language, their eyes are attracted by objects that have
similar names in the other language: they never switch off either language
entirely (Spivey and Marian 1999, 2003). Both phonological systems are
activated when producing cognates (Hermans et al. 2011; Friesen and
Jared 2011). Monolingual native speakers do not have this complex inter-
woven system, except in as much as it parallels the use of two dialects
by the same person or the developmental transition from one grammar
to another – universal bilingualism in the terms of Roeper (1999) or
Mehrsprachigkeit in those of Wandruszka (1971).
We will not review here other evidence for the inter-relationships
between languages in multi-competence, which will come out in many
guises in the following chapters. The integration continuum model used in
Cook (2003) was drawn as an aid for visualising the diversity and complex-
ity of the relationships between the languages, going along a continuum
from total separation through different levels of interconnection to total
integration (see Figure 1.1).

Separation Interconnection Integration
LA
LA
LA &
LB LB
LB
Figure 1.1 The integration continuum of possible relationships in multi-competence

(Cook 2003, p. 9)
At the separation pole of the continuum, the languages are completely

independent of each other, like Weinreich’s coordinate bilinguals
(Weinreich 1953); at the integration pole, they are totally integrated with
each other; in between come many possible degrees of interconnection.
The two poles are ideals that could never actually exist; all L2 users are
somewhere on the continuum in between. And of course different aspects
of language may be located at different points of the continuum; the
lexicon may be well integrated, as we have seen, syntax perhaps less so.
The continuum is not static but dynamic, moving constantly as the influ-
ence of particular languages waxes and wanes, variously through attrition
and transfer between some or all of the languages in multi-competence,
and through activation of language mode in speech. But the direction of
movement may be in either direction; an L2 user’s multi-competence may
separate the languages more over time or integrate them more.
This implies then that individuals vary greatly in the relationships
between the languages of their multi-competence, depending on many
factors. To progress, SLA research needs to get away from generalisations
that apply to all L2 users. Rather than a single common system for L2 users,
there may be many possible systems, unlike the relatively uniformity of
monolinguals. Putting learner groups to one side, Cook (2009b) defined
five groups of L2 users:
– ‘people using an L2 globally for a wide range of functions’,

– ‘people using an L2 internationally for specific functions’,
– ‘people using an L2 within a larger community’,
– ‘people historically from a particular community (re-) acquiring its
language as an L2’,
– ‘people using an L2 with spouses, siblings or friends’.
Such a scheme begins to cover the varieties of L2 users and uses. In

particular it distinguishes between research with L2 learners and with L2
users; L2 learners in classrooms are subject to a different set of influences
and language input from L2 users, inevitably reflecting decisions made by
language teachers and educational systems about teaching goals, methods
and techniques; they are more the product of their circumstances than
specimens of ‘pure’ language learning.

10 VIVIAN COOK
Multi-competence has affinities with other views within SLA research

that treat the languages of the L2 user within a single over-arching system.
For example, Dynamic Systems Theory (DST) captures the flexibility and
interconnectedness of language systems that are never static over both
short and long periods of time; any description is a single frame taken
from a continuous movie, as described by de Bot in Chapter 6 (this volume).
Recently the idea of Dominant Language Constellation (DLC) has been
proposed by Aronin (2006, p. 145) (see Chapter 7): ‘the group of the most
important languages for a particular individual, enabling as a whole unit,
the person to act in a multilingual environment and to meet all his/her
needs’. This conceptualises the relationship between the languages of
multi-competence in the individual and in the community as a constella-
tion of inner circle languages, orbited by the languages of the linguistics
repertoire, surrounded by an Oort cloud of languages the person is merely
aware of to some degree. In practical terms the number of languages in a
DLC seem to be about three, with the others coming into play in particular
circumstances. DLC is one useful way of looking at multilingualism from a
multi-competence perspective.
The concept of transfer, alias cross-linguistic influence, also takes on a
different meaning in multi-competence (Cook, to appear): the L1 part of
the system may influence the L2 part, the L2 may influence the L1, the L3
may influence the L2, and so on for all the relationships detailed in Jarvis
and Pavlenko (2009). Attrition of the first language too comes to have a
different meaning (Schmid 2011); rather than the metaphor of the first
language being lost or ground down, multi-competence balances itself in
a kind of eco-system: one language’s gain is another language’s loss.
Multi-competence is not a frozen state but a continuous interaction
between the different languages in the community and the individual.
The consequences of Premise 1 extend beyond the areas of bilingual-
ism and SLA research to all of linguistics. For example, historically the
norms for native speakers have often been established from L2 users,
whether Voice Onset Time for Japanese based on Japanese in the USA, as
pointed out by Kato (2004), Hopi grammar established from a native
speaker living in New York (Whorf 1940/1956), or Greek path preference
based on Greeks living in the USA (Papafragou et al. 2008). The language
informants called on by linguists or the participants in experiments
may respond differently from monolingual native speakers because of
the influence of their other languages. People who know more than one
language are suspect informants on their first language: ‘the judgments
about English of Bloomfield, Halliday or Chomsky are not trustworthy,
except where they are supported by evidence from “pure” monolin-
guals’ (Cook 2002a, p. 23), by virtue of the influence of the second
language that each of these linguists knows. For these reasons, multi-
competence research has often dealt with speakers with minimal or
maximal knowledge and use of another language, not with polarised

monolingual and balanced bilingual speakers, as we see in Chapter 2. In

multilingualism research, the argument has been taken further; much
SLA research is using participants who know more than two languages
and hence fails to distinguish the nature of L2 learning from L3/Ln
learning; ‘the control for this variable is often poor, inadequate, if not
lacking altogether’ (de Angelis 2007, p. 6). At the very least this suggests
that all participants and informants in language-related research need to
be described in terms of their full language backgrounds, in linguistics
as much as in second language acquisition research.
Premise 2 Multi-competence does not depend on the monolingual

native speaker
A native speaker is typically considered to be ‘a person who has spoken a
certain language since early childhood’ (McArthur 1992, p. 682). The cru-
cial elements in this definition are: the speaker acquired language as an
infant, and has spoken it continuously throughout life. Clearly this state
cannot be achieved by any L2 user, with the exception of early childhood
bilinguals. An element that seldom emerges overtly in the definition is
that a native speaker is assumed to speak only one language: ‘From
Saussure to Chomsky “homo monolinguis” is posited as the man who
uses language’ (Illich and Sanders 1988, p. 52). The native speaker in
question is also tacitly assumed to speak a status version of the language,
in terms of British English, a Received Pronunciation accent actually
found in a small minority of native speakers, using ‘standard’ grammar,
not the you/yous distinction found in Geordie or the multiple negation
I never did nothing to no-one nowhere found in many regional and historical
varieties of English other than the present-day ‘standard’. Indeed some
Japanese students of English apparently think of the native speaker as
male, white and, hopefully, handsome (Takamishi 2013).
The concept of native speaker is then highly simplified, excluding all but
the monolingual speakers of a standard form of the language. The native
speaker is seen as knowing the abstract institutional form of the language
meant when people say ‘I love the French language’ (Cook 2010). This
reflects the prestige form of the language that grammar books and diction-
aries of the language are based on, usually associated with the pretensions
of a nation state, English as the language of England, Chinese as the
language of China. Hence the term ‘native speaker’ invokes aspects of
national or group identity; it is a small step further to assume that a native
speaker of French has to be born in France, not Burkina Faso, another step
to assume that they come from Paris, not Marseilles.
Undoubtedly most commonsense thinking about second language
acquisition by both the general public and linguists takes this idealised
monolingual native speaker to be the goal of second language acquisition:
‘Relative to native speaker’s linguistic competence, learners’ interlan-
guage is deficient by definition’ (Kasper and Kellerman 1997, p. 5) or the

12 VIVIAN COOK
book title Incomplete Acquisition in Bilingualism (Montrul 2008). Many bilin-

guals, teachers, psychologists and SLA researchers believe that the more
your speech resembles the native speaker’s the greater the achievement:
they too adopt the monolingual view of bilingualism (Grosjean 2008).
Conversely you have failed if other people can detect your foreign accent,
spelling mistakes, etc. The native speaker is the touchstone against which
all other speakers are tested – ‘an idealised monolingual native speaker,
who is held to be the ultimate yardstick of linguistic success’ (Ortega 2009,
p. 140). Yet it is logically impossible for any L2 learner to achieve native
speaker status according to the definition given above: you can’t change
your early childhood experiences. Hyltenstam and Abrahamsson (2003),
for instance, suggest ‘absolute native-like command of an L2 may in fact
never be possible for any learner’. How could it be? Why should it be?
Ducklings grow into ducks; cygnets into swans: there is little point in
lamenting the deficiencies of the duck compared to the swan. As the APA
Guidelines (2010) say:
Bias may occur when the writer uses one group (usually the writer’s own
group) as the standard against which others are evaluated . . . Authors
should recognize that differences arising from racial/ethnic comparisons
do not imply deficits.
But this is exactly the bias that occurs when research from the monolin-
gual perspective takes the native speaker group as the norm rather than
the L2 user. As Mauranen (2012, p. 4) points out, ‘monolingualism is
neither the typical condition nor the gold standard’.
Looking at multi-competence as ‘the overall system of a mind or a
community that uses more than one language’ implies that it exists in its
own right, not as an ancillary to the systems in monolingual minds or
communities. L2 users are unique users of multiple languages, not pale
imitations of native speakers. The presence of the first language in second
language acquisition makes the whole language system different from
that of a monolingual, affecting both the second language, the first and
any others in the system. Birdsong (2005, p. 320) claims ‘Neither of the two
languages of a bilingual can be expected to resemble that of a native
monolingual. Accordingly, non-nativelike performance is not necessarily
indicative of compromised language learning abilities.’
Premise 2 is then a declaration of independence for the L2 user from the
monolingual native speaker and the mono-language community. To quote
François Grosjean, the bilingual is ‘a specific and fully competent speaker/
hearer who has developed a communicative competence that is equal, but
different in nature, to that of the monolingual’ (Grosjean 1994, p. 1657).
There may indeed be languages that do not have living native speakers, as
in the revival of Hebrew in Israel in the twentieth century, or Miami-
Illinois, taught as an ancestral language to children by parents who do
not speak it natively (Hirata-Edds and Peter, Chapter 15, this volume).

There is inevitably a danger of emphasising the advantages of bilingual-

ism, the disadvantages of monolingualism and the like. However, which-
ever direction the comparison faces, it is ultimately a question of
difference, not superiority or deficit, of independence for both monolin-
gual and bilingual.
The history of modern SLA research starts with the independent gram-
mars assumption that L1 children have their own language systems at
different ages, independent of adult grammars (e.g. Klima and Bellugi
1966). The version of this assumption that dominated SLA research for
many years was ‘interlanguage’ – ‘the utterances which are produced
when the learner attempts to say sentences of a TL [target language]’
(Selinker 1972). This quotation shows that interlanguage had not cut
the umbilical cord between SLA research and native speakers; learners’
sentences were still seen as approximations to those of native speakers
through the construct of target language. Spolsky summarised this as
‘Native Speaker Target condition . . . Second language learner language
approximates native speaker language’ (Spolsky 1989, p. 34). Perhaps the
bulk of classical SLA research ever since has described and explained
approximation to the monolingual native speaker.
A few researchers have nevertheless insisted all along on the L2 user’s
independence, as in the Comparative Fallacy of comparing L2 users with
native speakers (Bley-Vroman 1983). Yet many still imply covertly that
the goal of learning a second language is to speak like a native. For
example a highly thought-of paper by Abrahamsson and Hyltenstam
(2009) is entitled ‘Age of onset and nativelikeness in a second language’:
the important question about age of acquisition is taken to be whether it
helps learners become like native speakers, not like successful L2 users;
the measure of success is ‘perceived nativelikeness’ of L2 Swedish
speakers’ accents in terms of Stockholm speech. It seems hard to counter
the assumption implicit in so much SLA research that linguistic difference
is linguistic deficit, even if this has been denied by Labov (1969) for
American Black English, Boas (1920/1940) and Sapir (1921) for non-
European languages and Bernstein (1971) for working-class dialects. The
monolingual perspective in much research is not acknowledged, let alone
justified.
Apart from clinical linguistics, no area of language study starts from
the assumption that its speakers are deficient in terms of some other
group, the ‘monolingual bias’ described in Cook (1997); it is the fallacy
that Plato drew attention to of dividing the people in the world in two,
one group consisting of Greeks, the other of non-Greeks (Plato c. 363 BC,
262c–d). When L2 users differ in pronunciation or syntax from monolin-
gual native speakers, this does not mean they are wrong – as L2 users in
their own right. A foreign accent does not necessarily show lack of profi-
ciency, only difference from the ideal native speaker, and may often be a
sign of national, ethnic or dialect identity. If it is acceptable for your

14 VIVIAN COOK
English to show you come from Glasgow, Texas or Sydney, why is the
same not true if you come from Tokyo, Berlin or Santiago?
The advocacy of Premise 2 coincided with a movement in applied
linguistics to question the power of the native speaker. Phillipson (1992)
argued that the role of the native speaker in language teaching was a
form of linguistic imperialism. Ex-colonial countries assert their power
by claiming to own their national language. When a language becomes
supercentral or hypercentral (de Swaan 2002), it is a great asset to a
country in political and economic terms. For example the view that native
speakers based in England should write the textbooks for teaching English
around the world was a godsend for many British publishers. People in
different countries nevertheless have the right to use a language as they
see fit; they are not tied down by the wishes of other countries. Some may
indeed choose a native model to aspire to, others may decide that a local
variety, a form of English Lingua Franca or an L2 user variety are more
appropriate. The same argument has been used by British jazz critic Stuart
Nicholson (2005) to argue for the independence of European jazz from
American jazz as a form of music in its own right, not be judged by its
Americanness.
Inevitably this line of thinking leads to questioning the appropriate-
ness of the native speaker model for language teaching (Llurda 2005). It
seems a universal, almost instinctive, assumption that native speakers
make better language teachers because of the authenticity of their lan-
guage and their immersion in the native culture: generations of expat
English teachers have made their living out of this. But, if native speaker
language and culture is no longer the target, there is no absolute virtue in
being a native speaker, whether linguistically or culturally. Non-native
speaker teachers may make better role models for the students because
they have travelled the same route as them and are living exemplars of
successful L2 users able to handle two languages at the same time. Indeed
Brown (2013) suggests that ‘L2 performance be assessed by multi-
competent speakers of the L2’: the appropriate people to judge L2 users
must themselves be L2 users.
This is not to say that the change in attitudes towards non-native
speaker teachers has improved their job prospects. In 2015 the
University of Essex website proclaimed ‘All languages are taught by native
or bi-lingual teachers’, the University of East Anglia ‘The majority of our
undergraduate students are taught by native French, Spanish or Japanese
speaking lecturers and tutors’, the Modern Language Centre in Kings
College London ‘All teaching staff are native speakers of the language
they teach’: native speaker teachers are alive and well and teaching at
English universities.
If the monolingual native speaker is no longer the only true owner of
a language, SLA research needs to investigate L2 learning and L2 use as
distinctive properties of L2 users. Research questions need to be couched

in terms that refer to L2 users as independent people rather than subordi-

nating them, implicitly or explicitly, to native speakers; SLA research is
not a matter of investigating the shortfall of L2 users compared to natives
but of seeing them as themselves. While lip-service is now often paid to
the multi-competence perspective in SLA research questions, the native
speaker is still a ghost in the machine in terms of research methodology,
whether grammaticality judgments, error analysis, speech processing
or obligatory occurrences, all of which typically involve ‘monolingual
bias’ through implicit comparison with natives (Cook 1997), a point devel-
oped by Vaid and Meuter in this volume (Chapter 4). This is not to say
that SLA research cannot borrow techniques derived from other areas,
such as fMRI, L1 developmental scales and so on, provided they are used
for its own purposes, not as a way of putting down the L2 user. Comparison
of one group of language users to another can sometimes be an effective
research tool – among many others; we can learn from the variety of
language users what is special about each. But the uniqueness of L2 users
will be forever hidden if they are always described in terms of native
speakers, rather like trying to fit a quart into a pint bottle.
From a multi-competence perspective, classical SLA research has estab-
lished very little about L2 users themselves, only vast quantities of infor-
mation about their differences and similarities compared to native
speakers. Some classical research can be reinterpreted by stripping away
the native speaker element and discovering what it may show about
L2 users in their own right. However, as we see in Chapter 2, multi-
competence has posed its own research questions and utilises research
methods that do not invoke the native speaker.
This premise has perhaps the most implications for language teaching.
We have already seen that the role of the native speaker teacher needs
reassessing. But so do the very goals of language teaching and those of
many language students (Cook 2007). If the purpose of language teaching
is not to speak like a native speaker, then just what is it? For practical
purposes much second language use may be with non-native speakers,
for example the use of English in the tourist industry regardless of the
traveller’s first language or the use of English by air traffic controllers
everywhere in the world. And indeed, if both the first and second
languages are so inextricably woven into everything that an L2 user does,
we need to rethink the widespread view in language teaching that the
students’ first language has no role in classroom learning.
Premise 3 Multi-competence affects the whole mind, i.e. all language

and cognitive systems, rather than language alone
As the concept of multi-competence developed, it became apparent
that it concerned the whole mind of the L2 user, not just language
and so was linked to wider cognitive processes and concepts. One
theoretical slant in contemporary acquisition studies is called by

16 VIVIAN COOK
Chomsky (2013) the ‘no-language’ position, typified by ‘language is

entirely grounded in a constellation of cognitive capacities that
each – taken separately – has other functions as well’ (Enfield 2010).
In other words, language is a mere artefact of other cognitive processes.
The alternative position is that language is a unique part of the human
mind with its own characteristics that cannot be explained solely in
terms of other ‘cognitive capacities’. Essentially this is another skirm-
ish in the territorial battle between linguists and psychologists that has
been waged for many a generation. The multi-competence perspective
can presumably be adopted equally by those who see language as an
independent cognitive system, mostly linguists, and those who see it as
an interaction of other cognitive systems, mostly psychologists. The
import of this premise depends more on one’s overall theoretical
orientation than do the other two premises, which in a sense apply to
any research with second languages.
Premise 3 extends multi-competence to cover any aspect of the L2 user’s
mind that may be connected to their multilingualism. In one view, the
linguistic and conceptual systems are partitioned from each other and
do not contribute to each other’s development or use. Cognition consists
of a set of unvarying universal concepts ‘essentially available prior to
experience’ (Chomsky 1991, p. 29) and ‘assumed to be both fundamental
and universal in the semantic organisation of language’ (Talmy 2007,
p. 81). While languages vary in principled ways, concepts are constant,
thus rejecting the linguistic relativity hypothesis that language affects
concepts: How could it if concepts are invariable?
The recent wave of research has, however, shown that some version
of linguistic relativity is tenable and empirically supported, even if still
hotly disputed. In the 2000s multi-competence research started to ask, if
speakers of different languages think differently, how do L2 users who
know more than one language think (Cook 2002b; Cook and Bassetti
2011)? Going beyond the cross-linguistic comparison of most linguistic
relativity research, the purpose was to see whether speakers of two lan-
guages think differently from monolingual native speakers and how an
individual L2 user deals with such conceptual differences. Linguistic rela-
tivity could be tested, not just by comparing the thinking of monolinguals
across languages, but by comparing L2 users with monolinguals. If there
are indeed cognitive differences between monolingual native speakers
and L2 users, the obvious conclusion is that this is an effect of the other
language or languages they have acquired.
The cognitive effects of L2 learning might be general for any learning
of other languages regardless of language, such as the benefits of greater
metalinguistic awareness (Bialystok 1991) or reasoning (Han and Ginsburg
2001). Peal and Lambert (1962, p. 20) summed up some fifty years ago:
‘Intellectually [the bilingual child’s] experience with two language systems
seems to have left him with a mental flexibility, a superiority in concept

formation, a more diversified set of mental abilities.’ Language teaching

has indeed often maintained that learning another language helps chil-
dren to think better, particularly classical languages. Boris Johnson, the
Mayor of London in the 2010s, claims ‘Latin and Greek are great intellec-
tual disciplines, forcing young minds to think in a logical and analytical
way’, a claim that probably has more to do with the teaching methods
employed than the languages concerned.
Or these effects might appear only for relationships between specific
first and second languages rather than second language learning in gen-
eral (Bassetti and Cook 2011). If you speak Japanese, which calls both the
leg and the foot ashi, what happens when you learn English, in which there
are different words leg and foot?
In the view reflected in say Levelt et al. (1999) and Slobin (1991),
lexical concepts underlie speech production: speech to Levelt involves
connecting lexicalised concepts to the mental lexicon, to Slobin select-
ing concepts that can be expressed in a particular language, ‘thinking
for speaking’. The concepts they are interested in are those that are
intimately tied to language production, not those which lie behind it.
Concepts lie on a continuum between lexicalised concepts, which
are necessarily embedded in language, and non-lexicalised concepts,
which are minimally linked to language itself.
As the following chapters show, many L2 researchers have used multi-
competence in the context of the thinking for speaking approach (Han
and Cadierno 2010; Pavlenko 2011). It is not clear that this is actually
linguistic relativity. In one sense it is tautologous to say that lexicalised
concepts affect the language we use. To explore the effects of language on
non-language-related concepts not involved in the selection of concepts
to be expressed, we need on the one hand to minimise the amount of
language involved in the task, on the other to examine aspects of lan-
guage such as word order that have purely syntactic meaning rather than
such language-related semantic notions as gender and motion. The test of
linguistic relativity is whether L2 users differ in concepts that are not
explicitly used in speaking (Cook 2015), an issue developed by de Groot
(Chapter 12).
Finally the conundrum that still has to be solved is the meaning of
the word language, as we see in several chapters that follow, taken up
by Singleton (Chapter 24). In most SLA research language is taken as a
primitive term, so obvious in meaning that it needs no discussion.
However its meaning changes from one theory to another and one
context to another and indeed does not necessarily translate into
other languages (Wierzbicka 2014). For example the difference between
English two-way language and speech and French three-way langue,
langage and parole (de Saussure 1916/1976) has often puzzled English-
speaking linguists. The division between the ‘no language’ position and
generative linguistics arguably reduces to a dispute over the meaning

18 VIVIAN COOK
of language and is unresolvable without agreement on a mutually accep-

table meaning of language.
In Cook (2010) I argued that this issue lies at the heart of the disputes
that have riven SLA research. Jackendoff (2002) distinguishes:
(i) language as a countable whole (‘The French language’)

(ii) language as a set of abstract formal expressions (S→aSb) and
(iii) language that instantiates the rules of (ii) into meaningful
representations.
Cook (2010) tried to sketch six meanings of language that are encountered
in language research, namely:
– Lang1 human representation system – i.e. ‘humans possess language’,

– Lang2 an abstract external entity – ‘The French language’,
– Lang3 a set of actual or potential sentences – ‘the language of Molière’,
– Lang4 the possession of a community – ‘the language of French people’,
– Lang5 the knowledge in the mind of an individual – ‘I know French’,
– Lang6 a form of action – ‘I sentence you to twenty years imprisonment’.
We always need to be aware how the many concepts of ‘language’ are

used in second language and multilingualism research. Tomasello (2003,
p. 7), for example, claims that ‘the principles and structures whose
existence is difficult to explain without universal grammar . . . are
theory-internal affairs and simply do not exist in usage-based theories of
language – full stop’. Statements about Lang3 are indeed not statements
about Lang5. But the reverse also applies: statements about mental Lang5
make no claims about corpus-based Lang3, an issue discussed at least
since Chomsky (1957). Neither do statements about mental Lang5 always
connect to the institutional Lang2: individuals do not necessarily use the
rules, etc., of the formally described language in their everyday speech
and writing. Indeed, any individual person only commands a fraction of
the lexical and grammatical resources of the standard Lang2 in their
mental Lang5.
The Lang2 sense of an abstract entity – objective knowledge in Popper’s
World 3 of abstract ideas (Popper 1972, p. 159) – is central to any discussion
of second language learning since we need a label for the overall object,
whether the French language, Chinese or whatever. Dictionaries and
grammar books are usually descriptions of Lang2; since the eighteenth
century, national identity has been seen as a matter of Lang2 (Anderson
1983), leading to the rise of the assumption that monolingualism is the
norm (Yildiz 2012). But this does not necessarily have implications for
the organisation of language in the mind of the individual, called Lang5:
the mental representation of language is a complex system with all sorts
of internal and external relationships; it may be quite arbitrary to divide a
bilingual system into separate areas, modules and subsystems, that can
be called languages in the plural; multi-competence is a complex overall

system that may be indivisible into languages. Language in sense Lang2 is

countable, in sense Lang5 uncountable. So, while we can say someone
speaks two languages, English and French, in sense Lang2, this does not
mean that they necessarily have two discrete languages in their mind in
sense Lang5. It is easy to slip into reifying objects called languages in the
mind which relate together like the coloured balls in Bohr models of
the atom. Lang5 is, however, more like an amorphous indeterminate entity
from quantum theory than a clearly defined object.
So it is a moot point whether one can count Lang5 languages in the mind:
Can the multi-competent system be divided into separate sub-systems
labelled languages without destroying or denying the whole system, as
discussed by de Bot (Chapter 6, this volume)? And the same for the com-
munity: the Polish of the London Polish community is not the Polish of
Warsaw, the Cantonese of the Newcastle community is not that of Hong
Kong and so on. The languages in the multilingual community are so
tied in to each other that it becomes arbitrary to separate them. What is
needed is more like a DST of constantly changing relationships in which
some grouping of elements may be arbitrarily and temporarily called a
language in sense Lang5 without accepting that it is a discrete ‘language’
object. Just like the phoneme, so Lang5 language can be seen as a conveni-
ent shorthand for a complex of features.
1.5 Conclusion
The idea of multi-competence then raises a number of issues for all the
disciplines dealing with the acquisition and use of language. These con-
cern fundamental assumptions about language, about community and
identity and about research methodology. They have yielded a generation
of research on different lines, to be described in Chapter 2. We should
never forget that the L2 user is not an outsider lurking on the outskirts of
society but is in the main throng of humanity today. It is for instance
notoriously hard to find participants for language research anywhere in
the world who are ‘pure’ monolinguals untouched by other languages,
particularly English. Human beings have a potential, not for acquiring
one language as in Chomskyan discourse, but for acquiring more than
one language. Monolingualism is a problem in that it is a restriction on
human potential and a partial account of what makes a human being:
‘Speaking another language is quite simply the minimal and primary
condition for being alive’ (Kristeva 2007).
In terms of numbers, it is no more possible to count how many L2 users
there are in the world than to count how many monolinguals there are.
Claims that the majority of the human race are now L2 users (Clyne 1997)
are impossible to substantiate, however plausible. The sheer numbers
of speakers of languages that are not the central languages of their

20 VIVIAN COOK
countries go some way to show how many L2 users there must be. To take
some arbitrary figures related to English:
– 42.6 per cent of people in California speak another language than

English in the home (US Census Bureau 2010);
– over 90 per cent of secondary school children in Europe learn English as
a second language, 73 per cent of primary school children (EACE 2014);
– 546,000 Polish speakers are living in the UK (Office for National
Statistics 2012), 119,528 in Ireland (An Phrı́omh-Oifig Staidrimh 2012);
– out of the 3.3 million UK residents with a main language other than
English, 79 per cent could speak English very well or well (Office for
National Statistics 2012).
Research into second language acquisition, bilingualism and multilin-

gualism is not a fringe discipline but concerns central aspects of
human life for individuals and for communities in the twenty-first
century. Indeed the argument for the normalcy of L2 users means
that linguistics has to decide whether basing itself on the question
‘What constitutes knowledge of language?’ (Chomsky 1991) is a sig-
nificant distortion of the real question ‘What constitutes knowledge of
languages?’
Note
1. The web page Images of SLA (http://homepage.ntlworld.com/vivian.c/SLA/

SLAbookcovers.htm) illustrates this from the covers of bilingualism
books, the main alternatives to two heads being diagrams of chaos and
pictures of homunculi, i.e. one head inside another.
References
Abrahamsson, N. and Hyltenstam, K. 2009. ‘Age of onset and nativelike-

ness in a second language: listener perception versus linguistic scru-
tiny’, Language Learning 59, 249–306.
An Phrı́omh-Oifig Staidrimh/Central Statistics Office 2012. This is Ireland:
Highlights from Census 2011. Dublin: Stationery Office.
Anderson, B. 1983. Imagined Communities. New York: Verso.
APA Guidelines 2010. Publication Manual of the American Psychological
Association, sixth edition. Washington, DC: American Psychological
Association.
Aronin, L. 2006. ‘Dominant language constellations: an approach to multi-
lingualism studies’, in M. Ó Laoire (ed.), Multilingualism in Educational
Settings, pp. 140–159. Hohengehren: Schneider Publications.
Baker, P. and Eversley, J. 2000. Multilingual Capital. London: Battlebridge.

Bassetti, B. and Cook, V. J. 2011. ‘Relating language and cognition: the

second language user’, in V. J. Cook and B. Bassetti (eds),
pp. 143–190.
Bernstein, B. 1971. Class, Codes and Control, vol. 1. London: Routledge.
Bialystok, E. 1991. ‘Metalinguistic dimensions of bilingual language profi-
ciency’, in E. Bialystok (ed.), Language Processing in Bilingual Children,
pp. 113–140. Cambridge University Press.
Bialystok, E. 2009. ‘The Good, the Bad and the Indifferent’, Bilingualism:
Language and Cognition 12 (1), 3–11.
Birdsong, D. 2005. ‘Nativelikeness and non-nativelikeness in L2A
research’, IRAL 43, 319–328.
Bley-Vroman, R. 1983. ‘The comparative fallacy in interlanguage studies:
the case of systematicity’, Language Learning 33, 1–17.
Block, D. 2007. Second Language Identities. London: Continuum.
Bloomfield, L. 1933. Language. New York: Holt, Rinehart and Winston.
Boas, F. 1920/1940. ‘The methods of ethnology’, American Anthropologist 22,
311–322. Reprinted in F. Boas 1940, Race, Language and Culture, pp. 281–289.
New York: Free Press.
Brown, A. 2013. ‘Multi-competence and second language assessment’,
Language Assessment Quarterly 10 (2), 219–235.
Brutt-Griffler, J. 2002. World English: A Study of its Development. Clevedon:
Multilingual Matters.
Cabrelli Amaro, J., Flynn, S. and Rothman, J. (eds) 2012. ‘Introduction’, in
J. Cabrelli Amaro, S. Flynn and J. Rothman (eds), Third Language Acquisition
in Adulthood, pp. 1–6. Amsterdam: John Benjamins.
Canagarajah, S. 2007. ‘Lingua Franca English, multilingual communities
and language acquisition’, Modern Language Journal 91, 923–939.
Chomsky, N. 1957. Syntactic Structures. The Hague: Mouton.
Chomsky, N. 1991. ‘Linguistics and cognitive science: problems and
mysteries’, in A. Kasher (ed.), The Chomskyan Turn, pp. 26–53. Oxford:
Blackwell.
Chomsky, N. 2013. ‘Problems of projection’, Lingua 18, 1–35.
Chomsky, N. and Lasnik, H. 1993. ‘Principles and parameters theory’, in
J. Jacobs, A. von Stechow, W. Sternefeld and T. Vennemann (eds), Syntax:
An International Handbook of Contemporary Research, pp. 506–569. Berlin:
De Gruyter.
Clyne, M. 1997. ‘Multilingualism’, in F. Coulmas (ed.), The Handbook of
Sociolinguistics, pp. 301–314. Oxford: Blackwell.
Cook, V. J. 1991. ‘The poverty-of-the-stimulus argument and multi-
competence’, Second Language Research 7 (2), 103–117.
Cook, V. J. 1997. ‘Monolingual bias in second language acquisition
research’, Revista Canaria de Estudios Ingleses 34, 35–50.
Cook, V. J. 2002a. ‘Background to the L2 user’, in V. J. Cook (ed.), Portraits of
the L2 User, pp. 1–28. Clevedon: Multilingual Matters.

22 VIVIAN COOK
Cook, V. J. 2002b. Bilingual cognition. Opening talk at a colloquium at

EUROSLA Basel. http://homepage.ntlworld.com/vivian.c/Writings/
Papers/BilCogEUROSLA2002.htm
Cook, V. J. (ed.) 2003. Effects of the L2 on the L1. Clevedon: Multilingual
Matters.
Cook, V. J. 2007. ‘The goals of ELT: reproducing native-speakers or promot-
ing multi-competence among second language users?’ in J. Cummins
and C. Davison (eds), International Handbook of English Language Teaching,
pp. 237–248. Dordrecht: Kluwer.
Cook, V. J. 2009a. ‘Multilingual Universal Grammar as the norm’, in I.
Leung (ed.), Third Language Acquisition and Universal Grammar, pp. 55–70.
Bristol: Multilingual Matters.
Cook, V. J. 2009b. ‘Language user groups and language teaching’, in
V. J. Cook and Li Wei (eds), Contemporary Applied Linguistics, vol. 1:
Language Teaching and Learning, pp. 54–74. London: Continuum.
Cook, V. J. 2010. ‘Prolegomena to second language learning’, in
P. Seedhouse, S. Walsh and C. Jenks (eds), Conceptualising ‘Learning’ in
Applied Linguistics, pp. 6–22. Basingstoke: Palgrave Macmillan.
Cook, V. J. 2012. ‘Multi-competence’, in C. A. Chapelle (ed.), The Encyclopedia
of Applied Linguistics, pp. 3768–3774. New York: Wiley-Blackwell.
Cook, V. J. 2013. ‘The language of the street’, Applied Linguistics Review 4 (1),
43–81.
Cook, V. J. 2015. ‘Discussing the language and thought of motion in second
language speakers’, Modern Language Journal 99 (supplement), 154–164.
Cook, V. J., to appear. ‘The relationship between the languages of multi-
competence’, in R. Alonzo (ed.), Cross-linguistic Influence. Bristol:
Cook, V. J. and Bassetti, B. (eds) 2011. Language and Bilingual Cognition. New
York: Psychology Press.
Craik, F. and Lockhart, R. 1972. ‘Levels of processing: a framework for
memory research’, Journal of Verbal Learning and Verbal Behavior 11,
671–684.
De Angelis, G. 2007. Third or Additional Language Acquisition. Clevedon:
De Saussure, F. 1916/1976. Cours de Linguistique Générale. C. Bally and
A. Sechehaye (eds) 1916. Critical edition by T. de Maurio. Paris:
Payothèque, Payot.
De Swaan, A. 2002. Words of the World: The Global Language System.
Cambridge: Polity Press.
Doughty, C. and Long, M. 2003. Handbook of Second Language Acquisition.
Oxford: Blackwell.
Education, Audiovisual and Culture Agency (EACE) 2014. Key Data
on Teaching Languages in Europe. https://webgate.ec.europa.eu/fpfis/
mwikis/eurydice/index.php/Publications:Key_Data_on_Early_
Childhood_Education_and_Care_in_Europe_%E2%80%93_2014_Edition

Enfield, N. 2010. ‘Without social context?’ Science 329, 1600–1601.

Franceschini, R. 2011. ‘Multilingualism and multi-competence: a concep-
tual view’, Modern Language Journal 95 (3), 344–355.
Friesen, D. C. and Jared, D. 2011. ‘Cross-language phonological activation
of meaning: evidence from category verification’, Bilingualism: Language
and Cognition 15 (1), 145–156.
Green, D. W. 1998. ‘Mental control of the bilingual lexico-semantic
system’, Bilingualism: Language and Cognition 1, 67–81.
Grosjean, F. 1994. ‘Individual bilingualism’, in R. E. Asher (ed.), The
Encyclopaedia of Language and Linguistics, pp. 1656–1660. Oxford:
Pergamon Press.
Grosjean, F. 2008. Studying Bilinguals. Oxford University Press.
Hammerberg, B. 2010. ‘The languages of the multilingual: some concep-
tual and terminological issues’, IRAL 48, 91–104.
Han, A. Y. and Ginsburg, H. P. 2001. ‘Chinese and English mathematics
language: the relation between linguistic clarity and mathematics
performance’, Mathematical Thinking and Learning 3, 201–220.
Han, Z.-H. and Cadierno, T. (eds) 2010. Linguistic Relativity in SLA: Thinking for
Speaking. Bristol: Multilingual Matters.
Haugen, E. 1953. The Norwegian Language in America. University of
Pennsylvania Press.
Hermans, D., Ormel, E., van Besselaar, R. and van Hell, J. 2011. ‘Lexical
activation in bilinguals’ speech production is dynamic: how language
ambiguous words can affect cross-language activation’, Language and
Cognitive Processes 26 (10), 1687–1709.
Hyltenstam, K. and Abrahamsson, N. 2003. ‘Maturational constraints in
SLA’, in C. Doughty and M. Long (eds), The Handbook of Second Language
Acquisition, pp. 539–588. Oxford: Blackwell.
Illich, I. and Sanders, B. 1988. ABC: Alphabetisation of the Popular Mind.
Berkeley: North Point Press.
Jackendoff, R. 2002. Foundations of Language. Oxford University Press.
Jarvis, S. and Pavlenko, A. 2009. Crosslinguistic Influence in Language and
Cognition. Abingdon: Routledge.
Kasper, G. and Kellerman, E. (eds) 1997. Communication Strategies:
Psycholinguistic and Sociolinguistic Perspectives. London: Longman.
Kato, K. 2004. Second Language (L2) Segmental Speech Learning: Perception and
Production of L2 English by Japanese Native Speakers. PhD dissertation,
University of Essex, UK.
Klein, W. 1986. Second Language Acquisition. Cambridge University
Press.
Klima, E. and Bellugi, U. 1966. ‘Syntactic regularities in the speech of
children’, in J. Lyons and R. J. Wales (eds), Psycholinguistics Papers,
pp. 183–208. Edinburgh University Press.
Kristeva, J. 2007. ‘The other language or the condition of being alive’.
http://translation.hau.gr/telamon/files/Kristeva%27s%20speech.pdf

24 VIVIAN COOK
Kuhn, T. 1962. The Structure of Scientific Revolutions. University of Chicago

Press.
Labov, W. 1969. ‘The logic of non-standard English’, Georgetown Monographs
on Language and Linguistics 22, 1–31.
Lambert, W. 1990. ‘Persistent issues in bilingualism’, in B. Harley, P. Allen,
J. Cummins and M. Swain (eds), The Development of Second Language
Proficiency, pp. 201–218. Cambridge University Press.
Levelt, W., Roelofs, A. and Meyer, A. 1999. ‘A theory of lexical access in
speech production’, Behavioral and Brain Sciences 22, 1–75.
Lightbown, P. and Spada, N. 2006. How Languages are Learned. Oxford
University Press.
Llurda, E. (ed.) 2005. Non-Native Teachers. New York: Springer.
Mauranen, A. 2012. Exploring ELF. Cambridge University Press.
McArthur, T. 1992. Oxford Companion to the English Language. Oxford
University Press.
Montrul, S. 2008. Incomplete Acquisition in Bilingualism: Re-examining the Age
Factor. Amsterdam: John Benjamins.
Nicholson, S. 2005. Is Jazz Dead? (Or Has It Moved to a New Address). London:
Routledge.
Office for National Statistics 2012. 2011 Census. www.ons.gov.uk/ons/
index.html
Ortega, L. 2009. Understanding Second Language Acquisition. London: Hodder
Education.
Papafragou, A., Hulbert, J. and Trueswell, J. 2008. ‘Does language guide
event perception? Evidence from eye-movements’, Cognition 108,
155–184.
Pavlenko, A. 2005. Emotions and Multilingualism. Cambridge University Press.
Pavlenko, A. (ed.) 2011. Thinking and Speaking in Two Languages. Bristol:
Peal, E. and Lambert, W. 1962. ‘The relation of bilingualism to intelli-
gence’, Psychological Monographs 76, 1–23.
Phillipson, R. 1992. Linguistic Imperialism. Oxford University Press.
Plato, c. 363 BC. ‘The Statesman, 262a’, in Plato in Twelve Volumes, vol. 12,
translated by H. N. Fowler, 1921. London: William Heinemann.
Popper, K. 1972. Objective Knowledge: An Evolutionary Approach. Oxford
University Press.
Roeper, T. 1999. ‘Universal bilingualism’, Bilingualism: Language and
Cognition 2, 169–186.
Romaine, S. 1994. Bilingualism, second edition. Oxford: Blackwell.
Sapir, E. 1921. Language: An Introduction to the Study of Speech. Reprinted
(no date) by Harcourt Brace and Co, New York.
Schmid, M. S. 2011. Language Attrition. Cambridge University Press.
Selinker, L. 1972. ‘Interlanguage’, International Review of Applied Linguistics
10 (3), 209–231.

Skutnabb-Kangas, T. 1981. Bilingualism or Not: The Education of Minorities.

Clevedon: Multilingual Matters.
Slobin, D. I. 1991. ‘Learning to think for speaking: native language, cogni-
tion and rhetorical style’, Pragmatics 1 (1), 7–25.
Spivey, M. and Marian, V. 1999. ‘Cross talk between native and second
languages: partial activation of an irrelevant lexicon’, Psychological Science
10, 181–184.
Spivey, M. and Marian, V. 2003. ‘Competing activation in bilingual
language processing: within- and between-language competition’,
Bilingualism: Language and Cognition 6 (2), 97–115.
Spolsky, B. 1989. Conditions for Second Language Learning. Oxford University
Press.
Statistics Canada 2012. www12.statcan.gc.ca/census-recensement/2011/as-
sa/98-314-x/98-314-x2011001-eng.pdf
Stern, H. H. 1992. Issues and Options in Language Teaching. Oxford University
Press.
Takahashi, K. 2013. Language Learning, Gender and Desire. Bristol:
Talmy, L. 2007. ‘Lexical typologies’, in T. Shopen (ed.), Language Typology
and Syntactic Description, vol. III: Grammatical Categories and the Lexicon,
Tomasello, M. 2003. Constructing a Language. Cambridge, MA: Harvard
University Press.
US Census Bureau 2010. ‘Language use in the United States: 2007’, American
Community Survey Reports, ACS 12.
Wandruszka, M. 1971. Die Mehrsprachigkeit des Menschen. Munich: Piper.
Weinreich, U. 1953. Languages in Contact: Findings and Problems. The Hague:
Mouton.
Whorf, B. 1940/1956. ‘Science and linguistics’, in J. B. Carroll (ed.),
Language, Thought and Reality: Selected Writings of Benjamin Lee Whorf,
pp. 207–219. Cambridge, MA: MIT Press.
Wierzbicka, A. 2014. Imprisoned in English. Oxford University Press.
Yildiz, Y. 2012. Beyond the Mother Tongue: The Postmonolingual Condition. New
York: Fordham University Press.

2
Research questions
and methodology
of multi-competence
Goro Murahata, Yoshiko Murahata and Vivian Cook
2.1 Introduction
The aim of this chapter is twofold:
(i) to explore how linguistic multi-competence has raised research

questions that had rarely been asked previously, and
(ii) to introduce fundamental ideas about multi-competence-based
methodology and multi-competence-focused research questions.
Since its first mention in Cook (1991), the concept of multi-competence

has been developed in a number of studies, affecting an increasing number
of SLA researchers, teachers and students. Chapter 1 suggests that multi-
competence is primarily a different perspective or paradigm rather than a
disprovable model or theory. Hence it leads to new approaches and ques-
tions rather than experiments to prove or disprove its very existence. This
chapter does not try to survey all the possible research questions that
multi-competence may pose or indeed to review comprehensively all of
the relevant studies that have been conducted; further research can be
found in the bibliography of multi-competence we have compiled (this
volume). Rather it sketches out the kinds of question generated almost
uniquely by multi-competence and it provides a framework for exploring
second language acquisition from a multi-competence perspective.
One approach would be simply to critique existing second language
acquisition research using the three premises of multi-competence
outlined in Chapter 1. That is to say, research that was not carried out
with this point of view in mind can either be seen to be inadequate from
this perspective or can be reinterpreted from this perspective. Much of
the research into effects of age of start in second language acquisition
research is for instance presented in terms of the extent to which L2
users are defective compared with adult native speakers (Abrahamsson

Research questions and methodology of multi-competence 27
and Hyltenstam 2009). This research might in fact be reinterpreted as

establishing whether the unique competence of the L2 user is affected by
age, leaving the native speaker out of the equation. The approach here
tries to be positive by mostly detailing some of the research questions
that specifically arise from multi-competence, taking into account both
research overtly aimed at investigating multi-competence and research
that is compatible with multi-competence, rather than research that
needs to be stripped of its reliance on the native speaker before its rele-
vance can be assessed.
2.2 Research questions
Research Question 1: What are the effects of the languages

of multi-competence on each other?
As Premise 1 in Chapter 1 assumes, multi-competence covers the total
language system in a single mind or community. Interference, alias
transfer or cross-linguistic influence, is one way of conceptualising the
relationship between the languages; attrition is another. Classic main-
stream SLA research has always been interested in the influence of the
L1 on the L2, starting from the days of Contrastive Analysis and structur-
alist behaviourism that assumed that, as language was a matter of habit,
old L1 habits would affect new L2 habits. The influence of the L1 on the L2
has now been documented in countless PhD dissertations and articles
covering virtually every aspect of language from phonology to spelling.
The scope of multi-competence, however, extends to all the many relation-
ships between the languages in the L2 user’s mind described in Jarvis
and Pavlenko (2008), not just the L1 effects on the L2 (Cook, to appear).
When multi-competence consists of other languages than the first and
second, still more complex relationships come into play concerning the
influence of the L2 on the L3, the L3 on the L1 and so on (Cabrelli Amaro
et al. 2012). The overall research question which naturally arises is ‘What
are the effects of the languages of multi-competence on each other?’
A multi-competence perspective which regards L2 users’ competence
as a total language system necessarily attaches more importance to the
dynamic and permeable nature of linguistic competence in the mind
(de Bot, Lowie and Verspoor 2005; Ortega 2009; Li 2013; Pavlenko and
Malt 2011). Our linguistic competence as a whole may change as it
incorporates one or more linguistic systems, just like, say, the liquid
in a flask which changes, either slightly or drastically, as one or more
substances are put into it. In such a case, it is not easy to tell exactly
which substance affects the others in the flask. The chemical reaction of
multi-competence depends on how proficient the L2 user is in the second
language and how close or distant the first and the second languages are
linguistically. What is sure, however, is that the original liquid that

28 G O R O M U R A H ATA , YO S H I K O M U R A H ATA A N D V I V I A N C O O K
was in the flask at the start is no longer what it once was, nor can it ever
regain its original state.
Let us now look at some studies that explore the dynamic nature of L2
users’ multi-competence through more specific research questions (RQs).
The question that has been asked most often in multi-competence
research concerns the phenomenon of reverse transfer – the L2 effect on
the L1 – which had seldom been mentioned before multi-competence drew
attention to the bidirectional nature of L2 users’ linguistic competence,
though it was contained within Weinreich’s (1953) idea of transfer. L2
effects on the L1 and vice versa can manifest at various linguistic levels
such as phonology, lexicon (semantics), syntax or discourse. There is also
a logical distinction between macro-effects of bilingualism that will
happen to any L2 user regardless of their first and second languages
involved and micro-effects that pertain to the particular combination of
languages involved as L1 and L2 (Bassetti and Cook 2011). And indeed as
L3 and Ln.
RQ1a What impact does L2 learning have on the L2 users’ L1

phonology and vice versa?
Some incidental findings have supported reverse transfer in phonology.
Watson (1991) examined phonological features of French and English
monolinguals and bilinguals of different age groups. French and English
vary in, among others, Voice Onset Time (VOT) in plosive consonants
such as /t/. He found that while younger bilinguals behaved in a way
that was very similar to monolinguals, showing two independent sets of
contrasts, older/adult bilinguals behaved in a way that was distinct from
monolinguals of either language, showing a merging voicing feature.
According to Zampini and Green (2001), Spanish–English bilinguals
produce the L1 Spanish plosive /b/ with significantly longer VOTs than
Spanish monolinguals, and the L2 English /b/ and /p/ with significantly
longer VOTs than English monolinguals. Chang (2012) found effects of
learning L2 Korean on L1 English after only six weeks of instruction
while Chang and Mishler (2012) found that speakers of L1 Korean were
better at unreleased stops in L2 English than English native speakers.
This line of research nevertheless continued to interpret phonetic
acquisition as moving towards the target language, as seen in Chang
(2012, p. 252), ‘the longitudinal decline in ability to acquire L2 sounds
like a native speaker’, or in Harada (2007), who claimed that the VOTs
of Japanese–English bilingual children ‘are still [emphasis added]
intermediate, compared with the norms of the monolingual speakers’
(Harada 2007, p. 353). Wrembel (2011) has indeed extended this
research technique to the effects of the L1 and L2 on L3 VOTs. The
overall result for VOT is then that both the L1 and the L2 of L2 users
are different from monolinguals in both languages. At least in this area
of phonology, L2 users speak differently from monolinguals.

RQ1b Does the L2 affect the L1 semantics of L2 users?

In an early study directly examining multi-competence, Tokumaru (2002)
asked whether L2 English influences the semantics of English loanwords
in the Japanese of Japanese/English users. To monolingual Japanese speak-
ers, the English loanword bosu suggests ‘gangsters’ or ‘monkeys’ while to
monolingual English speakers boss suggests ‘office’ or ‘work’. As Japanese
learners acquire proficiency in English, they associate bosu less with
‘gangsters’ or ‘monkeys’ than monolinguals and more with ‘office’ or
‘work’. She concluded that ‘L2 Japanese users of English are different
from monolingual Japanese speakers . . . as a result of the influence from
their L2 knowledge of English’ (Tokumaru 2002, p. 405) in terms of the
association of L1 Japanese loanwords. Laufer (2003) also found L1 colloca-
tional knowledge among Russian–Hebrew bilinguals in Russian was
affected by their L2 Hebrew. Pavlenko (2003) reported L2 effects on L1 for
lexical borrowing, loan translation and semantic extension as well as
violation of tense and aspect, case-marking rules and prepositional choice
in L1 Russian in a story-telling task by Russian–English bilinguals. Wolff
and Ventura (2009) similarly found evidence for a reverse transfer effect
of L2 English on the meaning of causal verbs in L1 Russian. So far as
semantics is concerned, the reverse question of how the L2 affects the
L1 certainly shows a clear L2 on L1 effect in all these cases.
RQ1c What are the effects of the L2 on L2 users’ L1 syntax?

Cook, Iarossi, Stellakis and Tokumaru (2003) explored whether there
were differences in the L1 syntactic preference of monolinguals and the
L1 of L2 users who know another language, using Spanish, Greek and
Japanese L2 users of English and monolingual speakers of the same lan-
guages with minimal knowledge of English. Based on the Competition
Model methodology of investigating cues to the subject of the sentence,
advanced bilinguals behaved differently from novice bilinguals, for exam-
ple, with regard to animate cues and subject case cues. They concluded
that ‘this experiment has clearly shown that L2 users do not process the
sentences of their first language in the same way as monolingual native
speakers do’ (Cook et al. 2003, p. 212). Using eye-tracking, Dussias and
Sagarra (2007) examined how Spanish–English bilinguals process syntac-
tically ambiguous relative clauses such as The police arrested the brother of
the baby-sitter who had been ill for a while presented in their L1. Bilinguals
with extended exposure to English preferred the attachment of the
relative clause to the second noun more than bilinguals with limited
exposure to English, which they attributed to an L2 English effect on L1
Spanish syntactic processing.
Overall, then, L2 effects on L1 have been found at the phonological, lexical

and syntactic levels, and indeed also at the pragmatic levels (Blum-Kulka
1990; Heinz 2003; Krause-Ono 2004; Su 2010): our L1 is ‘more permeable

than traditionally thought’ (Li 2013, p. 158). Thus it is necessary for SLA
researchers to ‘examine which L2-influence processes take place in which
language areas, and what prompt particular changes’ (Pavlenko 2003, p. 58).
These three sub-questions represent the enterprise of looking for the
unique nature of the L2 user’s linguistic competence in the diverse areas
of language, and could be extended to many more seen in the other
papers in this volume. There are other, more speculative, research ques-
tions related to L2 influence on the L1 which might be explored in future
SLA research within the framework of multi-competence.
RQ1d How is L2-influenced L1 judged by the L1 community to which

L2 users belong?
As more evidence has validated L2 effects on L1, an issue arises about
the status of the native speaker. If the first language of multi-competent
L2 users is affected by the second language and is not the same as that
of L1 monolinguals, we need to investigate whether the L1 community
considers it deviant, wrong or strange (Brown and Gullberg 2012). Or
are its characteristics acceptable as unique in a positive sense? That is,
if the L1 Japanese of L2 English users is affected by English, and if the
Japanese–English bilingual speaks English-like Japanese, how is this
judged by monolingual Japanese speakers? In multi-competence commu-
nities like the Indian ‘Three language formula’ (Laitin 2000), similar ques-
tions can be asked just as in a single-language dominant community like
Japan, as seen above.
RQ1e Is there a threshold level of L2 knowledge that is necessary

to affect the L2 users’ L1?
Furthermore, it can be asked how much the L2 user has to know of the L2
before it affects their L1. Studies in this area typically use participants with
a wide range of proficiency in the L2, in order to reveal L2 effects compara-
tively easily among advanced L2 users. Brown and Gullberg (2012), how-
ever, find that L2 English has some effect on L1 Japanese at an earlier stage
of acquiring an L2 than expected. Longitudinal research on when an
individual L2 user may start to modify certain aspects of his/her L1 as he/
she acquires higher proficiency, and how it will change over time, would
therefore be interesting (Murahata 2010). This has repercussions for choice
of participants in any psycholinguistic experiment, as we see below: if
minimal exposure to a second language can affect the first language, any
so-called monolingual will have to be vetted for traces of the other lan-
guages they have encountered, even if only for a few years at school.
Research Question 2: Do L2 users think differently?

Premise 3 in Chapter 1 proclaims that multi-competence concerns the
whole mind of the L2 user, not just the language component of the

mind. It investigates the nature of both the linguistic aspects of the

bilingual mind and also the whole mind, which earlier SLA paradigms
have not been concerned with. Hence this immediately raises the
Sapir–Whorf hypothesis, alias linguistic relativity (Whorf 1940/1956),
which argues that the language we speak affects the way we think
about reality.
The first wave of research into linguistic relativity by Carroll and
Casagrande (1958) and Lenneberg and Robert (1956) in the 1950s
started to look at L2 effects on the mind long before multi-competence
emerged (see the account in Ervin-Tripp 2011). The 1990s resurgence
in empirical research into linguistic relativity (Gumperz and Levinson
1996; Levinson 1997; Lucy 1992) led to a view that the reason why
speakers of different languages pay attention to different aspects of
reality is indeed attributable to specific features of the languages they
speak (Agrillo and Roberson 2008; Boroditsky, Schmidt and Phillips
2003; Levinson 1997; Lucy 1992; Roberson, Davies and Davidoff 2000;
to cite a few). It is then quite reasonable to expect L2 users to think
differently from monolinguals because of the second language they
have alongside their first, and to ask whether they reshape particular
areas of their cognition as a consequence of learning another lan-
guage. Thus the scope of SLA research has expanded from language
constructs to encompass cognitive domains such as categorisation,
conceptualisation or perception of colour, objects, gender, space, emo-
tions, personhood, time, number and motion under such key terms as
‘concept merging’ (Ameel, Storms, Malt and Sloman 2005), ‘conceptual
shift/transfer’ (Jarvis and Pavlenko 2008; Pavlenko 2009), ‘bilingual
cognition’ (Cook and Bassetti 2011), ‘multi-cognition’ (Murahata
2010), ‘perceptual shift’ (Athanasopoulos, Damjanovic, Krajciova and
Sasaki 2011), or ‘cognitive restructuring’ (Athanasopoulos 2006).
RQ2a Do L2 users perceive colours differently from monolinguals?

Colour is a convenient research domain for testing how different segmen-
tation of the world affects cognition, since it presents a continuous reality
which is dissected variously by colour terms. When two colour terms in
one language are equivalent to a single term in another language, it can
be asked whether speakers of the former language are more sensitive to
the differences between the colours and recognise the two colours more
easily than speakers of the latter language. In terms of the ‘blue’ area,
Athanasopoulos (2009) showed Greek/English L2 users shift their percep-
tion of ble towards English blue, while their conceptualisation of ghalazio
is entirely different from either Greek monolingual ghalazio or English
blue. Athanasopoulos et al. (2011) found speakers of Japanese, another
language with two ‘blues’, ao and mizuiro, who used English frequently,
blurred the distinction. To sum up, bilinguals do perceive colours differ-
ently from monolinguals.

RQ2b Does L2 learning influence the way L2 users

categorise objects?
In the 2000s multi-competence studies could build on previous linguistic
relativity studies, originally conducted on monolingual speakers of differ-
ent languages and thus cross-linguistic rather than bilingual in nature.
Cook et al. (2006) replicated the shape-or-material matching experiments
with Japanese and English monolinguals by Imai and Gentner (1997)
with Japanese–English bilingual groups and English monolinguals; the
advanced Japanese–English bilingual group behaved more like the
English monolinguals than the low proficient Japanese–English bilingual
group. Athanasopoulos (2006) redid Lucy’s (1992) study of English
speakers and Yucatec speakers with Japanese–English bilinguals. He too
found growing sensitivity towards inanimate objects among advanced
Japanese–English bilinguals as they learn to pay attention to the number
of inanimate objects, which is not required in speaking Japanese. Based on
the study by Ji, Zhang and Nisbett (2004) which was conducted on English
monolinguals and Chinese–English bilinguals, Murahata (2012) revealed
L2 English effects on categorisation among Japanese–English bilinguals.
Presented categorical–thematic triads such as monkey–panda–banana,
where monkey and panda are categorically related and monkey and banana
are thematically related, Japanese–English bilinguals tended to relate
categorically related objects more strongly as they acquire higher profi-
ciency in English. People’s ways of categorisation are then influenced by
the languages they know.
RQ2c Does L2 grammatical gender marking affect cognition

in bilinguals?
In languages such as French, grammatical gender classifies nouns into
masculine or feminine in an ‘arbitrary’ fashion without regard to biologi-
cal gender, while English classifies them by their biological gender in
a ‘natural’ fashion. So does grammatical gender marking affect how
the speakers of those languages perceive objects? According to a cross-
linguistic study by Boroditsky et al. (2003), Spanish and German speakers
manifest grammatical gender effects on a memory task of object–proper
name pairs (e.g., apple–Patricia). They also relate certain adjectives to
objects based on grammatical gender of the language they speak: for
example, key is a feminine noun in Spanish llave and goes with adjectives
meaning golden, lovely and shiny; key is masculine in German Schlüssel and
so goes with adjectives meaning heavy, jagged and serrated. Forbes, Poulin-
Dubois, Rivero and Sera (2008), however, did not find clear L2 gramma-
tical gender effects in a task of assigning male or female voices to
pictured objects in their experiment conducted on French–English,
English–French, Spanish–English and English–Spanish bilinguals. On
the other hand, Bassetti (2011) revealed effects of L2 German gramma-
tical gender on L1 Italian users of German as an L2 in judging feminine

and masculine animals. Further research is clearly necessary since dif-

ferent tasks with varying combinations of languages yield different
results.
RQ2d Does the learning of a second language affect the way L2 users
perceive motion events?
An important influence on bilingual cognition research is the work
of Leonard Talmy, in particular his distinction between verb-framed
languages like Spanish which use the verb to express both motion
and path, Entra caminando ‘he enters walking’, and satellite-framed
languages like English which express path through a preposition or
particle, He walked in (Cook 2015; Talmy 2007). Does learning a second
language which modulates motions in a different way from the first
language affect the way L2 users perceive motions? Czechowska and
Ewert (2011) posed this research question with participants from
different language backgrounds, English and Polish. They found that
bilinguals’ reactions to motion differed from those of either monolin-
gual group, and that, while low proficient bilinguals had conceptual
shifts toward the L2, high proficient bilinguals restructured the con-
ceptual domain in a more complex way. Bylund and Jarvis (2011)
conducted a similar study on the impact of the differences in two
languages, Spanish and Swedish, in terms of the interaction of gram-
matical aspects and motion events. L2 Swedish users behaved
differently from Spanish monolinguals, showing clear L2 effects on
the L1 in this cognitive domain. There is now a large body of work
concerned with the perception of motion in L2 users, for example
Pavlenko (2011a), de Groot (2011), and Bylund and Athanasopoulos
(2015), broadly showing how L2 users perceive motion differently
from monolinguals.
RQ2e Are the brains of L2 users different from those of monolinguals?

For the past twenty years or so people have been discovering that
particular ways of thinking affect the actual physical structure of the
brain. We should not be surprised then to find Petitto et al. (2011,
p. 140) saying ‘exposure to greater than one language may alter neural
and language processing in ways that we suggest are advantageous to
language users’. On the one hand there are short-term effects: Kwok
et al. (2011) found a measurable brain effect after people had learnt
new colour words for one hour 48 minutes spread over three days.
Using a variety of brain measurement techniques, Osterhout et al.
(2008) claimed that ‘structural changes . . . can be observed after a
relatively short but intense instructional period’, namely first-year
French at a US university. Differences have been found in the brains
of bilingual babies in the first few months of life (Petitto et al. 2011),
showing that changes may be quick.

On the other hand, more long-term physical effects are seen in a

significant increase in proportional size in the corpus callosum (the
central area connecting right and left hemispheres) in language teachers
who had studied a second language for more than seven years (Coggins
et al. 2004). Della Rosa et al. (2013) found significant structural changes in
the brains of ten-year-old multilingual children over time. Overall, much
of the physical evidence shows the constant interaction between the
languages that is a feature of multi-competence. For instance Chinese
speakers still have Chinese active while reading English (Wu and Thierry
2010). Word retrieval largely activates the same area of the brain, to a
greater degree in the languages in which the participants are more
proficient (Halsband et al. 2002; Videsott et al. 2010). Overall, there
seems to be a number of physical differences between the brains of L2
users and monolinguals. While many of these are held to be beneficial to
the individual, as claimed by, say, Petitto et al. (2011) or Bialystok et al.
(2007), de Bruin, Treccani and Salla (2015) have pointed out that published
research may be publication-biased in favour of bilingual advantages for a
number of reasons.
RQ2f How are the new properties of L2 users’ cognition accepted

in the L1 community?
The speculative issue again arises over how the effects on L2 users’
thinking may appear to the community. Are the new properties of L2
users accepted as unique in a positive way, or are they considered as a
loss of L1 speakers’ language identity? For example, if Japanese–
English bilinguals do not distinguish the two blues ao and mizuiro in
the same way as Japanese monolinguals (Athanasopoulos et al. 2011),
do the Japanese feel they have lost some part of being Japanese? Is it
taken as a negative side of learning another language by the commu-
nity? We could also ask how L2 users themselves are perceived by
other L2 users. In the expanded area of this RQ2f, there is an interest-
ing study on multilingualism in the field of psychotherapy. The
research found that L2 user clients felt they were more connected to
a therapist, and therefore they felt more emotionally supportive, if
only they knew that the therapist spoke another language besides
their L1, whatever language it might be (Dewaele and Costa 2013).
In summary there is now a body of evidence that L2 users think differ-

ently from monolinguals. In addition to the areas mentioned here, evi-
dence is starting to come in that L2 users are more logical and less
emotional in their moral judgments in the second language than mono-
linguals (Costa et al. 2014), that young bilingual children are more crea-
tive at solving mathematical and non-mathematical problems than
monolinguals (Leikin 2012) and that L2 users have memory advantages
over L1 users between at least the ages of 40 and 70 (Ljungberg et al. 2013).

Research Question 3: Does multi-competence affect teaching

or classroom L2 learning?
Most of the first generation that called themselves SLA researchers
came either from bilingualism research or from language teaching.
Hence the implications of second language acquisition for other areas
have always been important; some SLA research is still part of the wing
of applied linguistics that tries to assist other areas concerned with
second languages. An honourable tradition that links SLA research
with language teaching runs from Lado (1964) through Wilkins
(1972) to Whong, Gil and Marsden (2013). The questions introduced
on pp. 27–32 above raise several issues about L2 teaching and
instructed L2 learning: role models of L2 users, the goals of learning a
second language, ideal L2 teachers and the ways of teaching a second
language. As these indicate, informing teachers about SLA research is
only one aspect of its application. Language teachers are seldom their
own masters but are in thrall to educational administrators, examina-
tion bodies, course-writers and indeed folk ideas about language learn-
ing that infuse parents, students and teachers alike.
One overall lesson from the multi-competence perspective is the
changed nature of the L2 user. Learning another language doesn’t just
give you a skill to add to those you already possess, such as keyboard-
ing or playing a musical instrument; it changes everything about you
from your first language to your brain. It is a metamorphosis into a
different person, the only parallel for which is the transforming effects
of literacy. Since the audiolingual method of the 1950s, through the
communicative and task-based methods, language teaching methodol-
ogy has taken the line that its goal is utilitarian – communication and
interaction with other people, rather than the development of the
individual through another language. The longstanding humanist tradi-
tion that language teaching educates the student in various ways to
become a ‘better’ person has gone by the board. Multi-competence
revives this tradition by insisting that the duty of language teachers
in general is not only to give students a communication skill but also to
help them become different people.
RQ3a Can L2 users accept successful L2 users as their role models

rather than native speakers?
A central purpose of the multi-competence perspective is to provide an
alternative role model to the native speaker (Premise 2 in Chapter 1).
So long as the native speaker is placed at the centre of L2 teaching and
is regarded as the standard model to emulate, the student’s goal is to
emulate the native speaker. Many students quite realistically see this
ultimate goal as almost impossible to achieve and so it gives them a
feeling of failure (see p. 12). L2 students, who are themselves complicit
with the monolingual perspective, as are their teachers, may think it

difficult to shift perspective and to become aware of their own unique-

ness. Instead successful L2 users can be role models presenting a more
realistic and attainable goal (Murahata 2006). The extent to which L2
users accept successful L2 users as their role models can be a good
research question; Timmis (2002) indeed showed the extent to which
students were affected by the monolingual perspective. We see below
that this raises the question of the respective values of native and non-
native speaker teachers. As always, language teaching may be torn
between its views of what is right and what the students think is right;
however valid the multi-competence perspective may be, the teacher
still has to accommodate to the students’ beliefs and stereotypes to
teach them successfully.
RQ3b What are the goals of L2 learning?

The idea embodied in Premise 3 (Chapter 1) that multi-competence affects
the whole mind has particular consequences for language teaching. The
goals of learning a second language can be both external and internal to
the student (Cook 2007); that is, they range from actual use of the target
language to internal growth of the student’s mind, or even brain. From
the multi-competence perspective, the goal of L2 learning is not only to
attain high proficiency in the target language as a skilled L2 user but also
to become a unique individual whose mind differs from monolinguals in
both languages. Aoki (2010) warns that English language teaching in
Japan has traditionally placed too much emphasis on proficiency develop-
ment, and he argues that the internal linguistic functions of language
related to cognition, thought and judgment should be acknowledged
more. Internal goals include better cultural attitudes and greater cognitive
flexibility. As we saw earlier, learning a second language can affect meta-
linguistic awareness (Bialystok 2001) and cognitive activities (Filippi et al.
2015), protect the brain against the symptoms of dementia (Bak and
Alladi 2014; Bialystok, Craik and Freedman 2007), and increase the grey
matter in the cortex (Kwok et al. 2011). Language teaching can now
point to direct benefits of L2 learning. The questions are how SLA
teachers/learners and the whole society will judge the notion that learning
an L2 aims at not only acquiring a high proficiency in the language but
also becoming ‘multi-cognitive L2 users’ (Murahata 2010).
RQ3c Who makes the best teacher of the second language – the
native or the non-native speaker?
Premise 2 of the multi-competence perspective, the independence of
the L2 user, casts a doubt over the long-term belief that ‘the ideal
teacher of English is the native speaker’ (Phillipson 1992, p. 185). In
practice, around the globe non-native English speaking teachers out-
number native English teachers. As Widdowson (1994, p. 387) points
out, it is the non-native teachers of English who are ‘in a better

position to know what is appropriate in contexts of language learning’,

while native English speaker teachers are ‘in a better position to know
what is appropriate in the context of language use’; from a multi-
competence perspective, ‘the developed multi-competence profile of
the teacher matches the developing multi-competent profile of the
learners’ (Brown 2013, p. 229). There has been considerable debate
on the advantages and disadvantages of native/non-native language
teachers and their contributions to second language acquisition
(Braine 2010; Llurda 2005), and further active discussion is still neces-
sary. Thus, the research question ‘Who is the ideal teacher of the L2?’ is
of great practical value in making the classroom an optimal learning
environment for L2 learners.
RQ3d How and when should the L1 be used in the L2 classroom?

In the mainstream language teaching tradition that goes back to the
nineteenth century, the use of the first language in the classroom
has been considered to be avoided or at least minimised based on a
common assumption that we acquire the L2 in the same way as the L1;
therefore, the L2 could best be taught and acquired exclusively by
using it for communication (Richards and Rogers 1986). However,
recently there have been keen calls to incorporate L2 users’ formerly
acquired cognitive and linguistic resources in the L2 language class-
room (Carson and Kashihara 2012; Ellis 2013; Matsumoto 2014; Taylor
and Snoddon 2013; among others). Carson and Kashihara (2012) men-
tion, as one of the effects of L1 use in the classroom, that allowing for
an increase in L1 use between students when working with old or new
materials or in groups helped promote production of the L2. When
non-native language teachers share the same first language with the
students, the L1 is always an option in the classroom, often a neces-
sity; one of the reasons for the reliance on the L2 is indeed the wide-
spread preference for expat native speaker teachers, who seldom
command the first language of their students. As multi-competence
considers two languages in one mind to be interwoven in a total
language system across many linguistic domains, rather than being
stored as two independent systems (Premises 1 and 3), the L2 class-
room should be a place where the L2 users can function with their full
language system. Teachers do not have to feel guilty when using the
first language. Students should not be punished when using the first
language, and be encouraged to use the L2 (Carson and Kashihara
2012). The principal aim is to use the first language positively in
teaching the second in order to produce L2 users capable of operating
with a holistic language system of languages as genuine L2 users
(Cook 2001). Thus the issue to be pursued in future SLA and teaching
research is how and when to use the first language rather than just
avoiding it.

2.3 The research methodology of multi-competence
Let us then see how the three premises of multi-competence relate to SLA
research methodology in principle, drawing on some of the points that
have already been made.
Premise 1 Multi-competence concerns the total system for all

languages (L1, L2, Ln) in a single mind or community and
their inter-relationships
Ideally any multi-competence-oriented SLA research should involve all
the languages in the mind of the individual or in the community, explicitly
or implicitly. Looking at the second language in isolation provides only
part of the picture; we cannot study, say, gender in the second language
without knowing about gender in the first. Knowledge of the first language
is often in a sense assumed; the grammar of the first language is described,
but not necessarily established, by the same techniques applied to the
second. From the multi-competence perspective, this is particularly
dangerous as it cannot be assumed that the first language of an L2 user
corresponds to the first language of a monolingual. Research that looks at
the L2 in isolation from the L1 reduces the study to quasi-L1 research,
ignoring the very factor that makes second language acquisition different –
the possession of a first language. The relationships between the
languages in multi-competence, whether transfer, attrition, or whatever,
are the core of second language acquisition research, its unique selling
proposition.
Premise 2 Multi-competence does not depend on the monolingual

native speaker
The ideal SLA research should therefore concentrate on the language
systems of L2 users in their own right, including first, second and any
other languages. It should not see any part of these L2 user systems as
an imperfect imitation of monolingual native speakers: the norm for
second language acquisition research is the L2 user, not the native speaker.
This does not rule out comparison of the monolingual and L2 user systems
as providing useful insights into the unique qualities of monolinguals and
L2 users. It does rule out any comparison that treats one as a defective
version of the other, whether seeing monolinguals as lacking qualities of
L2 users or L2 users as having defective versions of monolingual systems.
Native-speaker-centred monolingual perspectives on second language
acquisition miss the unique qualities of L2 users and how they develop:
comparison can only go so far. In a way, the multi-competence perspective
takes the ‘descriptive’ approach central to all linguistics and tries to
describe how L2 users are, rather than the ‘prescriptive’ approach, which
prescribes how L2 users should be.

Premise 3 Multi-competence affects the whole mind, i.e. all language

and cognitive systems, rather than language alone
The ideal multi-competence research sees language as one of the
complex systems in the mind, interacting with many other cognitive
systems. Thus it potentially has to take into account possible relation-
ships between language and other systems, such as memory systems,
conceptual structures or whatever. Research with, say, syntax may con-
nect to conceptualisation, research with learning strategies may con-
nect to motivation, etc. The research does not necessarily stop with the
language area involved but has implications for other areas of the
complex bilingual mind. This does not mean accepting the ‘no-lan-
guage’ position labelled by Chomsky (2013) that sees language as a
mere label for the intersection of multiple processes since language is
treated as a distinct system in its own right, albeit interacting with
many other systems.
On a strict interpretation of the three premises, classical SLA research
has accumulated considerable amounts of information about L2 users’
deficiencies vis-à-vis native speakers and very little about their unique
qualities, due to its use of research methods such as grammaticality
judgments, error recognition and correction, error analysis, fill-in-the-
blanks or obligatory occurrences (Larsen-Freeman and Long 1991) that
are inherently native speaker biased (Cook 1997).
The typical research design in multi-competence

Cook (2015) describes three idealised research designs for investigating
bilingual cognition.
(i) The four-language design. This examines a single property of language

across four languages, two of which possess it, two of which don’t, say
grammatical gender, and tests it across eight groups – four monolin-
gual and four bilingual. If all four bilingual groups show gender
assignment difference from monolinguals, this would show a
macro-effect of language learning, separate from specific language
effects.
(ii) The three-language design. This involves two first languages, one of
which shares a property such as grammatical gender with the target
L2, and tests two groups of L2 users, three groups of monolinguals. If
both bilingual groups assign gender the same way, this also suggests a
macro-effect, if less strongly.
(iii) The two-language design. This uses one L1 and one L2 which differ in
terms of the property, say having and not having grammatical
gender, and tests one group of L2 users, two groups of monolin-
guals. A good example is I-Ru Su’s (Chapter 14, this volume) use of
monolingual speakers of English and Chinese and bilingual

speakers of Chinese/English. This design can show micro-effects

from the interaction of a particular pair of languages but not a
macro-effect. This is perhaps the weakest in terms of generalisa-
bility since it can never be clear if the results are simply peculiar
to that pair of languages.
Alternative but complementary typologies of research designs are pre-

sented in Ortega (Chapter 3) and Vaid and Meuter (Chapter 4).
Till now a common research design in overtly multi-competence
research has been a variant on the two-language design involving the
comparison of people who speak a second language with those who
don’t, either in their first language or in terms of concepts. It is comparison
within speakers of a single language, not with native speakers nor with
other L2 users. The crucial factor is then to have at least two groups based
on proficiency in their L2 (Athanasopoulos 2006; Cook et al. 2006; to cite
two): one group of as near monolinguals as possible with least L2 knowl-
edge, sometimes called the minimal or low bilinguals, and an advanced
group called the maximal or high bilinguals, usually university students
of English.
The division into two groups is not then a straightforward choice
between monolinguals and bilinguals. On the one hand there are few
pure monolinguals since secondary schools now teach second lan-
guages all over the world, as indeed do many primary schools. It is
rare to find an ‘average’ person who has never been taught a second
language; any total monolinguals must be untypical as they have
evaded the normal educational requirements through isolation, depri-
vation or learning disability. It may well be that a minimal amount of
second language teaching doesn’t affect the L2 learner and that many
L2 learners end up still effectively monolingual. However, this needs
to be established empirically rather than assumed, particularly when
research such as that of Kwok et al. (2011) shows brain effects after
only 108 minutes of instruction and Chang (2012) shows phonological
L1 effects after six weeks. A fairer name for this group is then low
bilinguals rather than monolinguals, contrasting with high bilinguals.
This is not simply the native speaker norm sneaking in by the back
door under another name but a comparison group to see if L2 users
behave or respond like monolinguals.
It is also necessary to establish how well the L2 user group knows the
second language. So L2 users are required to take a test of language
proficiency as part of the experimental procedure, typically the Levels
Test of Nation (1990) used in Cook et al. (2006) or the Oxford Quick
Placement Test – usually the test published by Oxford University Press
(2001) used in Athanasopoulos and Kasai (2008) and Athanasopoulos
(Chapter 17, this volume) rather than that issued by the Oxford
University Language Centre (OULC 2015) found in Coventry et al. (2011).

Clearly at the moment these are based on native speaker norms rather
than those of L2 users.
This design is the standard for Research Question 1, which reveals the
effects of L2 learning on the first language of L2 users by comparing the
L2 users’ knowledge of the first language with that of monolinguals.
The design also executes Research Question 2 concerning the effects of
L2 learning on L2 users’ thinking by comparing the thinking of the L2
users with that of their monolingual peers.
Let us expand on some recent research that emphasises the cogni-
tive structure of the L2 user’s mind as a whole (Premise 3). Lucy (2011)
suggests two research approaches to the investigation of language
and thought: structure-centred approach and domain-centred. The
structure-centred approach starts with detailed analysis of the lan-
guages involved. When different linguistic forms in a specific area of
meaning are observed, then different interpretations of reality may be
revealed. The domain-centred approach starts with investigating how
various languages encode a certain domain of experienced reality such
as colour and space. Whichever approach may be taken, there are
three elements, that is, language, thought and reality, which should
be involved. Different words (language) to describe something (reality)
should go with different ways of looking at the part of reality
(thought). In particular, multi-competence research has often empha-
sised that a true test of the linguistic relativity hypothesis involves the
effects of language on non-language thought rather than, say, the
thinking-for-speaking approach of Slobin (1991), which is deliberately
confined to language-related areas (Cook 2015). Hence the aim is
for the cognitive task that people do to be as free from language as
possible, the better to study effects on non-language areas.
The tasks used in multi-competence differ little from those in classic
SLA and bilingualism research. What is different is the lack of involve-
ment of the native speaker to establish deficit, with the realisation that
comparison cannot reveal unique qualities of L2 users except inciden-
tally. Some techniques inherently rule themselves out. Grammaticality
judgments, for example, have been a standard SLA research technique
(e.g. Abrahamsson and Hyltenstam 2009; Kweon and Bley-Vroman
2011), though little employed in L1 acquisition research, despite some
worries whether such subjective self-assessment actually provides
access to the user’s linguistic competence. The answer to a grammati-
cality judgment is either right or wrong or on a scale between right and
wrong – but what are right and wrong? In multi-competence, rightness
would be conformity to a standard L2 user grammar, possible in prin-
ciple but in practice unachievable in our current state of knowledge.
Hence grammaticality judgments have almost inevitably to use native
speaker L1 grammar as a touchstone, whether based on linguists’

descriptions or on grammaticality judgments. Furthermore, since

multi-competence insists on the transformative potential of second
language learning, people who have learnt a second language have a
different metalinguistic awareness from monolinguals (Bialystok 2001).
L2 users will indubitably have different judgments from monolinguals
by virtue of the very fact of being L2 users. It is dangerous to use
grammaticality judgments in multi-competence research without
many caveats. The same applies inter alia to the research techniques
of elicited imitation and obligatory occurrences, where the native
speaker seems so embedded in the technique that the monolingual
perspective cannot be avoided.
Techniques that work with language data directly do not have the
same limitations. Learner-produced data can be analysed in their own
right to establish the grammars of L2 users rather than deviance from
the native, as in Klein and Perdue (1997). However this requires a form
of comparative grammar that does not force categories from the second
language on the first, for example such English-slanted statements
as such and such a language ‘lacks’ articles or inflections. The use of
experimental techniques such as reaction timing (Neubauer and
Clahsen 2009), eye-tracking (Cook et al., in progress) or EEG (Wu and
Thierry 2010) is not inherently native-speaker-biased provided any
comparisons with natives are based on difference, not deficit. A more
detailed critique of research techniques from the multi-competence
perspective can be found in Cook (1997).
2.4 Conclusion
In this chapter we have reviewed research questions within the frame-

work of multi-competence, which concern not only language but also
the whole minds of L2 users. They centred around L2 effects on L1,
L2 effects on thought/cognition, and the significance of the multi-
competence notion and its corollaries for teaching/learning an L2.
Then we introduced some fundamental beliefs on research methodol-
ogies and experimental tasks that would give the answers to those
research questions.
Research results have been accumulating to support the validity of the
multi-competence notion as a new perspective from which to view SLA
research. However, multi-competence research is still in its infancy and
its full implications are yet to be tested, for example which aspect of an
L2 language makes what kind of effects on which cognitive area. But it
shows a way towards getting a clearer portrait of the special properties
of L2 users, who after all now probably form a majority of human beings
in the twenty-first-century world.

References

tiny’, Language Learning 59, 249–306.
Agrillo, C. and Roberson, D. 2008. ‘Colour language and colour cognition:
Brown and Lenneberg revisited’, Visual Cognition 17 (3), 412–430.
Ameel, E., Storms, G., Malt, B. C. and Sloman, S. A. 2005. ‘How bilinguals
solve the naming problem’, Journal of Memory and Language 53, 60–80.
Aoki, S. 2010. ‘ “Kotoba no chikara” no ikusei ni tsuite omou’ [A thought on
nourishing ‘language literacy’], English Classwork 62 (2), 1.
Athanasopoulos, P. 2006. ‘Effects of grammatical representation of
number on cognition in bilinguals’, Bilingualism: Language and Cognition
9, 89–96.
Athanasopoulos, P. 2009. ‘Cognitive representation of colour in bilinguals:
the case of Greek blues’, Bilingualism: Language and Cognition 12, 83–95.
Athanasopoulos, P., Damjanovic, L., Krajciova, A. and Sasaki, M. 2011.
‘Representation of colour concepts in bilingual cognition: the case of
Japanese blues’, Bilingualism: Language and Cognition 14 (special issue),
9–17.
Athanasopoulos, P. and Kasai, C. 2008. ‘Language and thought in bilin-
guals: the case of grammatical number and nonverbal classification
preferences’, Applied Psycholinguistics 29, 105–123.
Bak, T. H. and Alladi, S. 2014. ‘Can being bilingual affect the onset of
dementia?’, Future Neurology 9 (2), 101–103.
Bassetti, B. 2011. ‘The grammatical and conceptual gender of animals in
second language users’, in V. J. Cook and B. Bassetti (eds), pp. 357–384.
second language user’, in V. J. Cook and B. Bassetti (eds), pp. 143–190.
Bialystok, E. 2001. Bilingualism in Development: Language, Literacy and
Cognition. Cambridge University Press.
Bialystok, E., Craik, F. I. M. and Freedman, M. 2007. ‘Bilingualism as a
protection against the onset of symptoms of dementia’, Neuropsychologia
45, 459–464.
Blum-Kulka, S. 1990. ‘You don’t touch lettuce with your fingers: parental
politeness in family discourse’, Journal of Pragmatics 14 (2), 259–288.
Boroditsky, L., Schmidt, L. and Phillips, W. 2003. ‘Sex, syntax, and seman-
tics’, in D. Gentner and S. Goldin-Meadow (eds), Language in Mind:
Advances in the Study of Language and Cognition, pp. 61–79. Cambridge,
MA: MIT Press.
Braine, G. 2010. Nonnative Speaker English Teachers: Research, Pedagogy, and
Professional Growth. New York and London: Routledge.
Language Assessment Quarterly 10, 219–235.

Brown, A. and Gullberg, M. 2012. ‘Multi-competence and native speaker

variation in clausal packaging in Japanese’, Second Language Research
28 (4), 415–442.
Bylund, E. and Athanasopoulos, P. (eds) 2015. ‘Cognitive SLA: the language
and thought of motion in second language speakers’, Modern Language
Journal 99, 1–13.
Bylund, E. and Jarvis, S. 2011. ‘L2 effects on L1 event conceptualization’,
Bilingualism: Language and Cognition 14, 47–59.
Cabrelli Amaro, J., Flynn, S. and Rothman, J. 2012. ‘Introduction’, in
J. Cabrelli Amaro, S. Flynn and J. Rothman (eds), Third Language
Acquisition in Adulthood, pp. 1–6. Amsterdam: John Benjamins.
Carroll, J. B. and Casagrande, J. B. 1958. ‘The function of language classifi-
cations in behavior’, in E. E. Maccoby, T. M. Newcombe and E. L. Hartley
(eds), Readings in Social Psychology, third edition, pp. 18–31. New York:
Holt, Rinehart and Winston.
Carson, E. and Kashihara, H. 2012. ‘Using the L1 in the L2 classroom: the
students speak’, The Language Teacher, 36 (4), 41–48.
Chang, C. B. 2012. ‘Rapid and multifaceted effects of second-language
learning on first-language speech production’, Journal of Phonetics 40 (2),
249–268.
Chang, C. B. and Mishler, A. 2012. ‘Evidence for language transfer leading
to a perceptual advantage for non-native listeners’, Journal of the Acoustical
Society of America 132 (4), 2700–2710.
Chomsky, N. 2013. ‘Problems of projection’, Lingua 130, 33–49.
Coggins, P. E., Kennedy, T. J. and Armstrong, T. A. 2004. ‘Bilingual corpus
callosum variability’, Brain and Language 89 (1), 69–75.
competence’, Second Language Research 7, 103–117.
Cook, V. J. 2001. ‘Using the first language in the classroom’, Canadian
Modern Language Review 57 (3), 402–423.
Cook, V. J. (ed.) 2003. Effects of the Second Language on the First. Clevedon,
Cook, V. J. 2007. ‘The goals of ELT: reproducing native-speakers or
promoting multi-competence among second language users?’, in
J. Cummins and C. Davison (eds), International Handbook of English
Language Teaching, pp. 237–248. New York: Springer.
Cook, V. J. 2015. ‘Discussing the language and thought of motion in
second language speakers’, in E. Bylund and P. Athanasopoulos (eds),
pp. 154–164.
Cook, V. J., to appear. ‘The relationship between the languages of
multi-competence’, in R. Alonzo (ed.), Cross-linguistic Influence. Bristol:

Cook, V. J., Alenazi, O., Alkadi, H. and Almutrafi, F., in progress. ‘Bilingual
street signs and eyetracking’.
Cook, V. J. and Bassetti, B. (eds) 2011. Language and Bilingual Cognition. Hove,
UK: Psychology Press.
Cook, V. J., Bassetti, B., Kasai, C., Sasaki, M. and Takahashi, J. A. 2006. ‘Do
bilinguals have different concepts? The case of shape and material in
Japanese L2 users of English’, International Journal of Bilingualism 10,
137–152.
Cook, V. J., Iarossi, E., Stellakis, N. and Tokumaru, Y. 2003. ‘Effects of the L2
on the syntactic processing of the L1’, in V. J. Cook (ed.), pp. 192–213.
Costa, A., Foucart, A., Hayakawa, S., Aparici, M., Apesteguia, J., Heafner, J.
and Keysar, B. 2014. ‘Your morals depend on language’, PLoS ONE 9 (4),
e94842.
Coventry, K. R., Guijarro-Fuentes, P. and Valdes, B. 2011. ‘Spatial language
and second language acquisition’, in V. J. Cook and B. Bassetti (eds),
pp. 263–286.
Czechowska, N. and Ewert, A. 2011. ‘Perception of motion by Polish–
English bilinguals’, in V. J. Cook and B. Bassetti (eds), pp. 287–314.
De Bot, K., Lowie, W. and Verspoor, M. 2005. Second Language Acquisition: An
Advanced Resource Book. London: Routledge.
De Bruin, A., Treccani, B. and Salla, S. D. 2015. ‘Cognitive advantage in
bilingualism: an example of publication bias?’, Psychological Science 26 (1),
99–107.
De Groot, A. M. B. 2011. Language and Cognition in Bilinguals and Multilinguals:
An Introduction. Hove, UK: Psychology Press.
Della Rosa, P. A., Videsott, G., Borsaa, V. M., Canini, M., Weekes, B. S.,
Franceschini, R. and Abutalebi, J. 2013. ‘A neural interactive location
for multilingual talent’, Cortex 49 (2), 605–608.
Dewaele, J.-M. and Costa, B. 2013. ‘Multilingual clients’ experience of
psychotherapy’, Language and Psychoanalysis 2 (2), 31–50.
Dussias, P. E. and Sagarra, N. 2007. ‘The effect of exposure on syntactic
parsing in Spanish–English bilinguals’, Bilingualism: Language and
Cognition 10 (1), 101–116.
Ellis, E. 2013. ‘The ESL teacher as plurilingual: an Australian perspective’,
TESOL Quarterly 47 (3), 447–471.
Ervin-Tripp, S. 2011. ‘Advances in the study of bilingualism: a personal
view’, in V. J. Cook and B. Bassetti (eds), pp. 219–228.
Filippi, R., Morris, J., Richardson, F. M., Bright, P., Thomas, M. S. C.,
Karmiloff-Smith, A. and Marian, V. 2015. ‘Bilingual children show an
advantage in controlling verbal interference during spoken language
comprehension’, Bilingualism: Language and Cognition 18 (3), 490–501.
Forbes, J. N., Poulin-Dubois, D., Rivero, M. R. and Sera, M. 2008.
‘Grammatical gender affects bilinguals’ conceptual gender: implications
for linguistic relativity and decision making’, The Open Applied Linguistics
Journal 1, 68–76.

Gumperz, J. J. and Levinson, S. C. 1996. ‘Introduction: Linguistic relativity

re-examined’, in J. J. Gumperz and S. C. Levinson (eds), Rethinking
Linguistic Relativity, pp. 1–18. Cambridge University Press.
Halsband, U., Krause, B. J., Sipiläd, H., Teräs, M. and Laihinen, A. 2002. ‘PET
studies on the memory processing of word pairs in bilingual Finnish–
English subjects’, Behavioural Brain Research 132 (1), 147–157.
Harada, T. 2007. ‘The production of voice onset time (VOT) by English-
speaking children in a Japanese immersion program’, IRAL 45,
353–378.
Heinz, B. 2003. ‘Backchannel responses as strategic responses in bilingual
speakers’ conversations’, Journal of Pragmatics 35, 1113–1142.
Imai, M. and Gentner, D. 1997. ‘A cross-linguistic study of early word
meaning: universal ontology and linguistic influence’, Cognition 62,
169–200.
Cognition. New York: Routledge.
Ji, L.-J., Zhang, Z. and Nisbett, R. E. 2004. ‘Is it culture or is it language?
Examination of language effects in cross-cultural research on categor-
ization’, Journal of Personality and Social Psychology 87 (1), 57–65.
Klein, W. and Perdue, C. 1997. ‘The basic variety (or: couldn’t natural
languages be much simpler?)’, Second Language Research 13 (4), 301–347.
Krause-Ono, M. 2004. ‘Change in backchannelling behaviour: the influ-
ence from L2 to L1 on the use of backchannel cues’, Studies of Cognitive
Science (Society of Muroran Cognitive Science) 3, 51–81.
Kweon, S.-O. and Bley-Vroman, R. 2011. ‘Acquisition of the constraints on
wanna contraction by advanced second language learners: Universal
Grammar and imperfect knowledge’, Second Language Research 27 (2),
207–228.
Kwok, V., Niu, Z., Kay, P., Zhou, K., Jin, Z., So, K.-F. and Tan, L.-H. 2011.
‘Learning new color names produces rapid increase in gray matter in
the intact adult human cortex’, PNAS 108 (16), 6686–6688.
Lado, R. 1964. Language Teaching: A Scientific Approach. New York:
McGraw-Hill.
Laitin, D. D. 2000. ‘What is a language community?’ American Journal of
Political Science 44 (1), 142–145.
Larsen-Freeman, D. and Long, M. L. 1991. An Introduction to Second Language
Acquisition Research. New York: Longman.
Laufer, B. 2003. ‘The influence of L2 on L1 collocational knowledge and
on L1 lexical diversity in free written expression’, in V. J. Cook (ed.),
pp. 19–31.
Leikin, M. 2012. ‘The effect of bilingualism on creativity: developmental
and educational perspectives’, International Journal of Bilingualism 17 (4),
431–447.
Lenneberg, E. H. and Robert, J. M. 1956. ‘The language of experience: a
study in methodology’, Perceptual and Motor Skills 12, 375–382.

Levinson, S. C. 1997. ‘Language and cognition: the cognitive consequences

of spatial description in Guugu Yimithirr’, Journal of Linguistic Anthropology
7 (1), 98–131.
Li, P. 2013. ‘Successive language acquisition’, in F. Grosjean and P. Li (eds),
The Psycholinguistics of Bilingualism, pp. 145–167. West Sussex, UK: Wiley-
Blackwell.
Ljungberg, J. K., Hansson, P., Andres, P., Josefsson, M. and Nilsson, L.-G.
2013. ‘A longitudinal study of memory advantages in bilinguals’, PLoS
ONE 8 (9), e73029.
Llurda, E. (ed.) 2005. Non-Native Teachers. New York: Springer.
Lucy, J. 1992. Grammatical Categories and Cognition: A Case Study of the Linguistic
Relativity Hypothesis. Cambridge University Press.
Lucy, J. 2011. ‘Language and cognition: the view from anthropology’, in
V. J. Cook and B. Bassetti (eds), pp. 43–68.
Matsumoto, Y. 2014. ‘Beyond the logic of “all or nothing” – Optimal use of
L1 and L2 in the classroom’. Paper read at the 43rd Annual Conference
of The Kyushu Academic Society of English Language Education, Oita
University, Oita, Japan.
Murahata, G. 2010. ‘Multi-cognition in L2 users: more evidence from an
object categorization task by Japanese elementary school children’,
Research Reports of Kochi University 59, 131–146.
Murahata, Y. 2006. ‘What do we learn from NNEST-related issues?
Some implications for TEFL in Japan’, The Language Teacher 30 (6), 3–7.
Murahata, Y. 2012. A Study of the Effects of English Learning on Cognition in
Japanese Learners of English from a Multi-competence Perspective. PhD disserta-
tion, Newcastle University.
Nation, P. 1990. Teaching and Learning Vocabulary. New York: Newbury
House/Harper Row.
Neubauer, K. and Clahsen, H. 2009. ‘Decomposition of inflected words in a
second language: an experimental study of German participles’, Studies
in Second Language Acquisition 31 (3), 403–435.
Arnold.
Osterhout, L., Poliakov, A., Inoue, K., McLaughlin, J., Valentine, G.,
Pitkanen, I., Frenck-Mestre, C. and Hirschensohn, J. 2008. ‘Second-lan-
guage learning and changes in the brain’, Journal of Neurolinguistics 21,
509–521.
Oxford University Language Centre. 2015. Oxford Placement Test. www.lang
.ox.ac.uk/courses/tst_placement_english.html
Pavlenko, A. 2003. ‘ “I feel clumsy speaking Russian”: L2 influence on L1 in
narratives of Russian L2 users of English’, in V. J. Cook (ed.), pp. 32–61.
Pavlenko, A. 2009. ‘Conceptual representation in the bilingual lexicon and
second language vocabulary learning’, in A. Pavlenko (ed.), The Bilingual
Mental Lexicon: Interdisciplinary Approaches, pp. 125–160. Clevedon, UK:

Pavlenko, A. (ed.) 2011a. Thinking and Speaking in Two Languages. Clevedon,

UK: Multilingual Matters.
Pavlenko, A. 2011b. ‘Overview of the field’, in A. Pavlenko (ed.),
pp. 237–257.
Pavlenko, A. and Malt, B. 2011. ‘Kitchen Russian: cross-linguistic differ-
ences and first-language object naming by Russian–English bilinguals’,
Petitto, L. A., Berens, M. S., Kovelman, I., Dubins, M. H., Jasinska, K. and
Shalinsky, M. 2011. ‘The “Perceptual Wedge Hypothesis” as the basis for
bilingual babies’ phonetic processing advantage: new insights from
fNIRS brain imaging’, Brain and Language 121 (2), 130–143.
Phillipson, R. 1992. Linguistic Imperialism. Oxford University Press
Richards, J. C. and Rogers, T. S. 1986. Approaches and Methods in Language
Teaching: A Description and Analysis. Cambridge University Press.
Roberson, D., Davies, I. and Davidoff, J. 2000. ‘Color categories are not
universal: replications and new evidence from a stone-age culture’,
Journal of Experimental Psychology: General 129, 369–398.
Slobin, D. I. 1991. ‘Learning to think for speaking: native language, cogni-
tion and rhetorical style’, Pragmatics 1 (1), 7–25.
Su, I-R. 2010. ‘Transfer of pragmatic competences: a bi-directional
perspective’, Modern Language Journal 94 (1), 87–102.
Talmy, L. 2007. ‘Lexical typologies’, in T. Shopen (ed.), Language Typology and
Syntactic Description, vol. III: Grammatical Categories and the Lexicon, second
edition, pp. 66–168. Cambridge University Press.
Taylor, S. K. and Snoddon, K. 2013. ‘Plurilingualism in TESOL: Promising
controversies’, TESOL Quarterly 47 (3), 439–445.
Timmis, I. 2002. ‘Non-native speaker norms and international English:
a classroom view’, English Language Teaching Journal 56 (3), 240–249.
Tokumaru, Y. 2002. ‘Cross-linguistic influences of L2 English on L1
Japanese in Japanese–English bilinguals’, ACTAS/Proceedings II Simposio
Internacional Bilingüismo, 399–406.
UCLES (University of Cambridge Local Examinations Syndicate) 2001. Quick
Placement Test. Oxford University Press.
Videsott, V., Herrnberger, B., Hoenig, K., Schilly, E., Grothe, J., Wiater, W.
and Spitzer, M. 2010. ‘Speaking in multiple languages: neural correlates
of language proficiency in multilingual word production’, Brain and
Language 113, 103–112.
Watson, I. 1991. ‘Phonological processing in two languages’, in E. Bialystok
(ed.), Language Processes in Bilingual Children, pp. 25–48. Cambridge
University Press.
Mouton.
Whong, M., Gil, K.-H. and Marsden, H. (eds) 2013. Universal Grammar and the
Second Language Classroom. Springer.

Whorf, B. L. 1940/1956. ‘Science and linguistics’, in J. B. Carroll (ed.),

Language, Thought and Reality: Selected Writings of Benjamin Lee Whorf,
Widdowson, H. G. 1994. ‘The ownership of English’, TESOL Quarterly 28,
377–389.
Wilkins, D. A. 1972. Linguistics in Language Teaching. London: Edward Arnold.
Wolff, P. and Ventura, T. 2009. ‘When Russians learn English: how the
semantics of causation may change’, Bilingualism: Language and Cognition
12, 153–176.
Wrembel, M. 2011. ‘Crosslinguistic influences in third language acquisi-
tion of Voice Onset Time’, in W.-S. Lee and E. Zee (eds), Proceedings of
the 17th International Congress of Phonetic Sciences, pp. 2157–2160. Hong
Kong: City University of Hong Kong.
Wu, Y. J. and Thierry, G. 2010. ‘Chinese–English bilinguals reading English
hear Chinese’, The Journal of Neuroscience 30 (22), 7646–7651.
Zampini, M. L. and Green, K. P. 2001. ‘The voicing contrast in English and
Spanish: the relationship between perception and production’, in
J. L. Nicol (ed.), One Mind, Two Languages: Bilingual Language Processing,
pp. 23–48. Oxford: Blackwell.

3
Multi-competence
in second language
acquisition: inroads into
the mainstream?
Lourdes Ortega
3.1 Introduction
In this chapter I reflect on the ways in which the perspective of multi-

competence has contributed to critically questioning research as we
know it in the field of second language acquisition (SLA). I endorse
Cook’s (2012) definition of multi-competence as “the knowledge of
more than one language in the same mind or the same community”
(p. 3768). In this chapter, however, I will concern myself with bilingual
or multilingual minds and will have little to say about bi/multilingual
communities.
I share several important tenets put forth by Cook (e.g. 2012, Chapter 1,
this volume) for multi-competence as a perspective:
– The learning of additional languages (L2s) later in life can and should be
reconstrued as the development of late bilingualism (Ortega 2013).
– Deficit orientations towards late bilingualism still plague many areas of
applied linguistics, and particularly SLA, and pose both a validity and an
ethics threat to disciplinary knowledge (Ortega 2014).
– A wholistic (or holistic) stance (Grosjean 1989) towards language,
language development, and language users is desirable.
– Language is constitutive of strategic, local actions (and agency), of our
being with others in the world (and identities), and of structured social
practices (and power) (Garcı́a and Li Wei 2014).
– Linguistic competencies and indeed language itself are dynamic and
they change at multiple time scales, including over the lifespan, as a
function of actual use (de Bot, Lowie, Thorne and Verspoor 2013; Five
Graces Group, Beckner et al. 2009).

Multi-competence in second language acquisition: inroads into the mainstream? 51
– Language is part of cognition and, as such, cognition and language

influence and affect each other (Bylund and Athanasopoulos 2014;
Langacker 2008; Pulvermüller 2013).
Finally, I also endorse the three premises of multi-competence put forth

by Cook in this volume (Chapter 1), reproduced here for ease of reading:
Premise 1 Multi-competence concerns the total system for all languages

(L1, L2, Ln) in a single mind or community and their inter-
relationships.
Premise 2 Multi-competence does not depend on the monolingual native
speaker.
Premise 3 Multi-competence affects the whole mind, i.e. all language and
cognitive systems, rather than language alone.
I will use Premises 1 and 2 as a heuristic for my examination in this

chapter. I have chosen to leave Premise 3 aside for several reasons.
One is that other authors in this handbook will be able to address how
multi-competence affects the whole mind in much greater depth than my
expertise would allow me to here. Another reason is that I take Premise 3
to be a matter of theoretical preference. That is, different researchers may
agree or disagree to investigate this issue, and to investigate it in different
ways at that. Good examples of SLA programs that align well with Premise
3 of multi-competence are research into how the experience of becoming
bilingual or using more than one language plays out when thinking for
speaking (Han and Cadierno 2010) and research on linguistic relativity
phenomena with bi/multilinguals (Bylund and Athanasopoulos 2014).
But, to my mind, the adoption of the bilingual perspective called for in
Premises 1 and 2 of multi-competence does not strictly necessitate a
commitment to Premise 3 on the relationship between language and
cognition in bilinguals and, vice versa, researchers may adhere to the
position that language and cognition affect each other and that bilingual-
ism affects this relationship without perhaps committing to the other
two premises. While I treasure the contributions of multi-competence to
the study of the relationship between language and cognition in late
bilinguals, it is Premises 1 and 2 of multi-competence that I find crucial
for advancing SLA as a whole field. More broadly, what I embrace as the
most important contribution and consequence of multi-competence is the
position that “someone who knows two or more languages is a different
person from a monolingual and so needs to be looked at in their own right
rather than as a deficient monolingual” (Cook 2012, p. 3768).
This chapter, then, is my attempt to chart the influence that multi-
competence has and has not had on SLA, using Premises 1 and 2 as
heuristics to gauge that influence. I highlight the benefits brought about
by the inroads that multi-competence has made into traditional SLA, while
also sketching areas where the impact has been more superficial than

52 LOURDES ORTEGA
substantive, a criticism that speaks of weaknesses endemic to the field.

Where possible, I also reflect on further disciplinary changes multi-
competence will likely (hopefully) help support in the future.
3.2 Capturing the total system, and the significance

of study designs
It is an obvious fact that the people investigated in SLA studies are by

definition multiple-language users engaged in the task of becoming bilin-
guals or multilinguals later in life. In attunement with this reality, the
multi-competence perspective entails as its first premise a focus on the
total system for all languages (L1, L2, Ln) in a single mind or community and their
inter-relationships. Yet, since the inception of the field and up to this date,
SLA researchers have mostly sought empirical evidence pertaining to
one language only, the target language exclusively, while they have kept
participants’ other languages outside the scope of investigation. This
research practice is in stark opposition to the practice favored in the
field of bi/multilingual first language acquisition, where within the same
study researchers routinely collect and analyze data for their participants’
“language pairs” or in the case of L3 acquisition, from the participants’
three languages (see, for example, studies reviewed in de Houwer 2009).
It is not, however, that SLA has missed the importance of accounting for
the languages (in the plural) of L2 users. Proof of this comes from the
sustained interest in the study of cross-linguistic influences, a vibrant
research domain within the field whose main goal is to understand the
roles that an individual’s first language(s) and other already known
languages (L3, Ln) play in the process of developing a new language at a
later time. A historical chronicle of such efforts can be easily tracked in the
eight book-length contributions by Gass and Selinker (1983), Kellerman
and Sharwood Smith (1986), Ringbom (1987, 2007), Odlin (1989), Jarvis and
Pavlenko (2008), de Angelis and Dewaele (2011), and Yu and Odlin (2015).
Yet the eschewing of participants’ other languages (L1, L3, Ln) is true even
in the majority of this research domain and, by comparison, SLA research
into cross-linguistic influence examining evidence beyond one single
language (the L2) is in short supply.
Indeed, many SLA studies have attempted to incorporate the other
languages of participants by resorting to strategies that ironically help to
circumvent the need for any empirical linguistic evidence outside the L2,
the target language under scrutiny. The application of such strategies
results in distinct study designs, which are summarized in Table 3.1.
Consideration of these study designs can help evaluate how well research
in SLA has to date addressed Premise 1 of multi-competence, or the
accounting of a bilingual’s total system of all languages and their inter-
relationships.

Table 3.1 Study designs for the investigation of cross-linguistic influence in SLA
Study Multiple-language
design L2 evidence? Comparisons/baselines? evidence? Total system?
Type 0 YES, a main group from the same L1 NO NO, data in target NO, each participant
background of interest language only contributes data in one
language only
Type 1 YES, a main group(s) from a given L1 YES, a native monolingual group as interpretive yardstick NO, data in target NO, each participant
background of interest of target language only contributes data in one
language only
Type 2 YES, a main group(s) from a given L1 YES, more than one native monolingual group as YES, data in target NO, each participant
background of interest interpretive yardstick of main participants’ target and L1 language and two or contributes data in one
more first languages language only
Type 3 YES, two or more main groups NO, L2-only group comparisons NO, data in target NO, each participant
performing in the same L2, sampled language only contributes data in one
from different L1s language only
Type 4 YES, two or more main groups YES, more than one base-line native monolingual group as YES, data in target NO, each participant
performing in the same L2, sampled interpretive yardsticks for each of the languages language and first contributes data in one
from different L1s language(s) language only
Type 5 YES, main groups share one language YES, more than one base-line native (monolingual and/or YES, data in L1, L2, Ln YES, L1, L2, Ln data elicited
(L2) at least and differ in one other bilingual) group as interpretive yardsticks (for from the same participants
language (L1, Ln) at least monolingual and/or bilingual competence) in the bilingual group(s)
54 LOURDES ORTEGA
The most basic design in cross-linguistic influence studies is Type 0,

where data are elicited from one group only and in the L2 or target
language only, and these single-language results are then interpreted by
recourse to certain descriptive (but empirically untested) L1–L2 differences
that are predicted to invite cross-linguistic influence. As a hypothetical
example (from Ortega 2009, p. 51), a researcher may recruit a group of L1
Spanish speakers learning English and interpret the observation that they
produce many non-inverted subject–verb questions (How I do this?) as
evidence that they are transferring the non-inversion rule for questions
from their L1. And while the researcher’s conclusion is consistent with
the descriptive L1–L2 differences, it would in fact be in error, as many
other studies with other groups whose L1s do have inversion have also
turned up un-inverted questions as typical of English at intermediate
stages of proficiency. Most contemporary cross-linguistic SLA studies
have therefore turned away from this design for its evidentiary thinness.
Types 1–4 represent more proactive strategies to account empirically for
the other language(s) of an L2 user, but only indirectly via inter-individual
comparisons. In Type 1 designs, L1 users of the target language under study
are sampled, so as to enable comparisons of the L1 performances of this
baseline group with the L2 performances of the main participant group(s).
This is perhaps the most widely employed design in cross-linguistic SLA
research. Early examples are two L2 English relative clause studies by
Schachter (1974) and Cook (1975). Schachter compared the amount of
relative clauses produced in a total of 200 essays written by four L2
English groups (with Persian, Arabic, Chinese, and Japanese as their L1)
against 50 essays written by college undergraduates in English L1. Cook
compared 52 adults learning English as a foreign language to 60 L1 English
elementary school children and 111 L1 English university students in
order to observe the relative cognitive ease or difficulty with which they
could comprehend relative clauses with single and double embeddings
(The dog that pushes the cat likes the horse, The dog that pushes the cat that sees the
man likes the horse). This kind of study design places a single language under
the investigative lens and employs a between-groups comparison of (one
or more than one group of) L2 users vis-à-vis (one or more than one group
of) L1 users of the given target language, with the latter group meant to
provide the interpretive yardstick for so-called ultimate and complete
linguistic competence. Type 2 studies adopt an improved variation of
this strategy, by including not one but at least two baseline L1 groups,
who are intended to yield normative evidence about not only the target
language but also the L1 of the L2 users under study. An example is Grüter,
Lieberman and Gualmini (2010), who conducted two parallel studies in a
between-group reverse learning design (the authors call it “bidirectional”)
that tapped English and Japanese performances, both as L2 as well as L1.
The main research question was whether the departing point for L2
acquisition would be the L1 or a universal default. The specific aspect of

grammar was negative scope; for example, a sentence like They don’t speak
English or German would be understood to mean not English and not German by
L1 English monolinguals and this is also the universal default interpreta-
tion that all children acquiring any L1 monolingually prefer until about
age 5; by contrast, not English . . . or maybe not German (I can’t remember) would
be the preferred L1 monolingual Japanese interpretation. In order to study
the problem more properly, the adult monolingual L1 facts were checked
empirically against L1 baselines for both L2s investigated (as it turned
out, the Japanese L1–English L2 group performed less well because they
preferred the interpretation available from their L1, which means they
did not revert to the universal default as a starting point; the English
L1–Japanese L2 group on the other hand performed better in learning a
new, non-default, interpretation). This design, involving as it does
between-groups comparisons of one (or more) L2 user groups against two
different L1 user groups (or more, if an Ln is involved), requires analysis
of more than one language within the same study. This is a welcome
strength. However, just as in the Type 1 design, the Type 2 design
(a) continues to turn to monolingual benchmarks of the participants’
linguistic competencies, in effect perpetuating monolingualism as the
golden norm, and (b) elicits only one language per group, even from the
L2 users, in effect erasing their bilingualism by treating participants who
are functional in more than one language as if they were monolinguals.
Study design Type 3 features a strategy whose virtues are discussed in
detail by Jarvis (2000) and Odlin (1989), although these researchers would
also encourage the addition of multiple L1 baselines, in effect favoring a
mixed Type 4 design. In Type 3, L2 users of two (or more) critically chosen
L1 backgrounds are sampled and then their respective L2 performances
are compared in order to link theoretically predicted L2 differences to
their L1 background differences. Examples of this design include White’s
(1985) oft-cited study of pro-drop, where she collected L2 English gramma-
ticality judgments about sentences such as the second one in “John is
greedy. Eats like a pig” (p. 51) from 54 speakers of Spanish, an L1 that
allows subjectless sentences (“Juan es un comilón. Traga como un cer-
dito”). In that study, as White explains, the L2 English performance of 19
French L1 speakers was also elicited and this group was considered a
control, “since French, like English, is not a pro-drop language” (p. 50).
Type 3 designs are welcomed in invoking direct L2–L2 comparisons as
interpretively rich and sufficient. But since only one of the participants’
languages, the L2, is inspected, once again the target language is investi-
gated in isolation.
Potentially, design Type 4 is also possible, by combining Types 1–3. This
happens when a study compares L2 groups from two or more L1 back-
grounds and supplements this comparison with a comparison against
several L1 groups which are meant to provide interpretive normative
yardsticks for all the languages (L1, L3, Ln) of the L2 user groups. Jarvis

56 LOURDES ORTEGA
(2000) offers an interesting illustration of such a mixed design. He com-

pared the lexical choices (e.g. a preference for car over van or truck, or take
over pick) in the L3 English narratives of a group of L1 Finnish–L2 Swedish
users and another group of L1 Swedish–L2 Finnish users; he chose as his
controls three groups of 22 age-matched participants each, who produced
the same narrative data in their English L1, Finnish L1, and Swedish L1
(all experimental and control groups of Finns and Swedes were high school
students from Finland; the L1 English controls were from high schools in
the Midwest in the United States).
In order for SLA research to address Premise 1 of multi-competence, that
is, for a design to account for the total system for all languages (L1, L2, Ln)
in a single mind and their inter-relationships, the kind of study which
is ideally needed would collect data from the same L2 user participants
across all their languages. This is study design Type 5 in Table 3.1. Only in
this way is the main participants’ bi/multilingualism invited into the
realm of what counts as relevant evidence to answer the field’s questions.
Thus, if in years to come the field of SLA as a whole persists in favoring
study design Types 1–4 in Table 3.1, it will continue to fail to address
Premise 1 of the multi-competence perspective.
3.3 Contributions of multi-competence to SLA

with Premise 1
An exemplary illustration of study design Type 5 is Brown and Gullberg

(2008). These researchers collected data in Japanese (L1) and English (L2)
by the same main participant groups, who were L2 users/learners of
English. At the same time, they also collected baseline monolingual
data from two groups, one comprising Japanese L1 users and one com-
prising English L1 users. It was precisely this design that enabled them
to discover slight but important differences in the way that the L2 users
synchronized their gestures and discourse in their two languages,
compared with the two L1 monolingual groups, with values in the
expression of motion events that fell in between the values of the
corresponding monolingual yardsticks for both Japanese (their L1) and
English (their L2). That is, and as predicted by the perspective of multi-
competence, these participants’ linguistic knowledge/use was more
than the sum of two monolingual competences. They exhibited bidirec-
tional cross-linguistic influence, from their Japanese L1 to their English
L2, and from their English L2 to their Japanese L1. As a result, they were
neither like English monolinguals in their English retellings nor like
Japanese monolinguals in their Japanese narrative retellings. In their
work, Brown and Gullberg explicitly make a link to the construct of
multi-competence (e.g. 2008, p. 246) and conclude that multi-competent
users will not look like mono-competent users of either their L1 or their

L2. That is, bi/multilinguals are not monolinguals, a tautological truism

that is customarily flouted in the field of SLA.1
In fact, by employing study design Type 5 repeatedly, Brown and
Gullberg (2008, 2012, 2013) have uncovered robust evidence of clear
multi-competence effects in a range of cross-linguistic influence phenom-
ena pertaining to the expression of motion events. Moreover, and surpris-
ingly under traditional SLA thinking, the accumulated work by Brown and
Gullberg (2008, 2012, 2013) has shown that such bidirectional influences
emerge even at early (intermediate) L2 proficiency levels, and even for
users with limited opportunities to participate in rich L2 usage because
they live in foreign language settings. Thus, it is important to underscore
that SLA researchers cannot consider themselves absolved from account-
ing for language interaction and cross-linguistic multidirectionality phe-
nomena, even when studying incipient levels of L2 development or when
dominance in the L1 remains intact.
Making a study accountable for the total system of all languages in a
single mind helps uncover findings about the nature of cross-linguistic
influence that are obscured when same-participant data are inspected
for only the L2. The first important finding is that transfer can travel in
multiple directions, as discussed in connection with Brown and Gullberg’s
(2008) work. Namely, although still the exception, when SLA researchers
have begun to implement a Type 5 design, they have turned up clear
evidence that all the languages of late bi/multilinguals influence one
another regardless of the sequence in which they were learned. Such
phenomena have been called reverse transfer and bidirectional transfer
in SLA (see Jarvis and Pavlenko 2008). Attention to these language interac-
tions was thrown into sharp relief with publications by Cenoz, Hufeisen
and Jessner (2001), Pavlenko and Jarvis (2002) and by Cook (2003), all three
adopting the multi-competence perspective explicitly. But the bidirection-
ality of cross-linguistic influence has been noted in SLA since the 1980s,
particularly by SLA phonologists who bothered to collect same-participant
L1 and L2 data from the L2 users they studied, such as Flege (1987). Indeed,
both Cook (1993, p. 2) and much later Jarvis and Pavlenko (2008, p. 12)
remind us that reverse transfer as an insight can be dated back to the work
by linguist Uriel Weinreich in the 1950s. Another interesting finding
revealed by a look at the total system of L2 users is that the timing of
learning is not the sole driving factor in language development, insofar as
the mature L1 is not singularly resilient to influence from other languages
that are learned later in life. Instead, the L1 can change over the lifespan as
a result of the experience of late bilingualism. This second finding
is associated with renewed interest in the study of L1 attrition among
late-starting L2 users (e.g. de Leeuw, Opitz and Lubińska 2013).
Clearly, in its empirical pursuit of Premise 1 and its concomitant promo-
tion of study design Type 5, the multi-competence perspective has made
two unique contributions into the SLA study of cross-linguistic influence

58 LOURDES ORTEGA
that are important to appreciate. First, it has enabled the replication and
extension to late-emerging bilinguals of the same well-established find-
ings among early or mature bilinguals put forth in the psycholinguistics of
bilingual processing (although this family of phenomena is often called
cross-language interactions in that field rather than bidirectional transfer;
see reviews by Kroll, Bogulski and McClain 2012, and van Hell and Tanner
2012). Second, and in doing so, the multi-competence perspective has
strengthened interdisciplinary ties, helping envision SLA as an integral
part of the study of bilingualism across the lifespan.
3.4 Benefits of addressing the total system that SLA

cannot pass up
To be clear, the adoption of study design Type 5 places extraordinary

demands on researchers. It requires bi/multilingual research teams with
expertise in all the languages of the L2 user participants and their baseline
groups, many more hours of coding and analysis, more sophisticated
statistical tests to deal with the design’s complexity, and many more
human participants to satisfy statistical assumptions and power in the
first place. When faced with these daunting requirements, traditional
SLA researchers will reasonably wonder just how advantageous, really,
Type 5 design is. All research is about a wise calibration of resources and
returns. Are the rewards worth such extraordinary demands?
Admittedly, it is possible to arrive at interesting findings about how the
languages of an L2 user affect one another solely by eliciting data in a
single language (the participants’ L2). This is more likely to happen, other
things being equal, provided the researchers adhere to the psycholinguis-
tic spirit of the total system (even though making themselves temporarily
empirically unaccountable to it in a given study), and if a tightly controlled
experiment is run with stimuli that induce particular L1–L2 interaction
effects. For example, de Zeeuw, Verhoeven and Schreuder (2012), on the
basis of a Type 1 design with Dutch data only, were able to conclude that
school-aged L1 Turkish–L2 Dutch users were not disadvantaged by their
smaller vocabulary size in Dutch, because processes of semantic activation
that went from the L1 to the L2 aided them in recognizing large-family-
sized Dutch words faster and more accurately, replicating the same facil-
itative effect for Dutch family size that was also seen in their age-matched
L1 monolingual Dutch peers with larger Dutch vocabularies.
So, beside the fact that design Type 5 heeds Premise 1 of the multi-
competence perspective to account for the total system for all languages
(L1, L2, Ln) in a single mind, are there any other reasons why in the future
SLA researchers might want to be more willing, within single studies, to
collect and analyze evidence from the same L2 users across their

languages? I can offer at least two, and both are relevant for all SLA
researchers, well beyond those who investigate cross-linguistic influence.
A first reason is that a deficit orientation unwittingly tints the results
for most language development phenomena under investigation when
only L2 evidence is collected, analyzed and interpreted. This is because
L2 users will inevitably look as if they have, know, do, or are less rather
than more, language-wise.
Researchers who study bilingual first language acquisition now recog-
nize that a false disadvantage will always emerge in data where only one
of the languages of bilingual children is measured (de Houwer 2009).
Indeed, they have come to the conclusion that, in order to avoid under-
estimation of bilingual abilities and misdiagnosis of deficits that do not
exist, measurements must not only capture performance in the multiple
languages of the bi/multilingual participants, but the evidence should also
be “totaled” or combined in some way, as it were. Perhaps the clearest
illustration can be found in the study of bilingual vocabulary develop-
ment, from Pearson, Fernández and Oller (1993) all the way to Core,
Hoff, Rumiche and Señor (2013). In this latter recent study, the researchers
confirmed that an inventory of the total vocabulary (i.e., the sum of all
word forms known in the two languages combined) provides the most
accurate estimate of size and rage of bilingual vocabulary development
for toddlers between 20 months and 33 months of age. This position is
thoroughly consistent with Premise 1 of the multi-competence perspec-
tive, which demands accounting for “the total system.”
Just as in the study of bilingual first language acquisition, which is
concerned with early bi/multilinguals, in the study of L2 acquisition too,
which is concerned with late-starting bi/multilinguals, conclusions based
on evidence from one language only (the L2) obscure changes and accom-
plishments in the linguistic systems of L2 users. Consider, for example, the
fact that SLA studies of L2 vocabulary size typically yield levels of L2
word knowledge that greatly lag behind the levels of single-language
word knowledge of L1 groups (see review in Schmitt 2014). Yet, when L2
users incorporate new L2 words into their overall conceptual and lexical
system – for example in higher-education contexts where the target
language is used to learn new content that is academic, specialized and
abstract – it may not be trivial at all to consider vocabulary size and depth
evidence in L2 and L1 from the same participants. Researchers would be
well advised to do so before negatively comparing their lexical prowess to
that of their L1 peers. Evidence in their L1 may be useful because indivi-
dual differences in L1 vocabulary knowledge are known to be very large
(Hoff 2003; Rowe et al. 2012). Further, vocabulary knowledge is likely to
interact bidirectionally across languages, from L1 to L2 (e.g. de Zeeuw et al.
2012; Kroll et al. 2012; van Hell and Tanner 2012) but also from L2 to L1,
for example, when efforts at learning new concepts and vocabulary in
an L2 might result in an expansion of one’s L1 vocabulary size and/or a

60 LOURDES ORTEGA
reorganization of one’s L1 vocabulary depth. Besides helping undo the

construction of deficit for international students who are L2 users, just
as in other areas of language (e.g., motion events, as in the work by Brown
and Gullberg 2008), new findings may emerge about vocabulary develop-
ment as a whole that could open up the pursuit of much more subtle
hypotheses about what learning new words means.
The practice of disregarding the total language knowledge of L2 users
and of isolating knowledge of the L2 as the only relevant object of inquiry
might help explain a host of suspect characterizations of semilingualism,
incompleteness, L1 attrition without L2 acquisition, and so on. These
proposals all share the fact that they assert that there are whole L2 user
populations who not develop real competence in any of their languages.
Some of these distorting characterizations have been later criticized and
discarded (see discussion about semilingualism in MacSwan 2000), but
plenty of them are still current in various language fields and are treated
as if they were neutral and plausible linguistic realities. I believe that if
more SLA researchers were to heed Premise 1 of multi-competence and
fewer went about their research business as usual, many of these proposals
would be much more quickly dispelled as simply wrong. As Brown and
Gullberg (2013) insist, “prudent research and pedagogy will not view an
individual’s multi-competent performance in one language in isolation
from their performance in the other language” (p. 491).
The second reason why it would be worthwhile for the field of SLA
as a whole to heed Premise 1 of multi-competence and make much
greater use of Type 5 designs is that this move would quickly place
SLA researchers in the position to contribute substantially to the study
of language dominance, an area that has recently attracted renewed
theoretical and methodological interest in bilingualism and psycholin-
guistics research (see Birdsong 2014, Silva-Corvalán and Treffers-Daller
in press, and Treffers-Daller 2011). Many, probably most, bilinguals
develop unequal skills in their languages, as a natural result of what
Grosjean (2008) calls the complementary principle: they use different
languages and combination of languages across different domains of
life and social networks. This being so, a central preoccupation among
bilingualism researchers is to understand the processes by which,
within the bilingual user’s total system, a language may weaken or
strengthen overall, with respect to the other(s). The place of these
concerns within the multi-competence perspective is questionable, as
the metaphor of dominance may serve to reify the idea of separate
languages that, instead of interacting dynamically, compete (Vivian
Cook, personal communication, July 18, 2014). Even more damaging is
the concern that the dominance literature might perpetuate what
Taylor (2011, p. 260) calls the hydraulic metaphor of bilingualism, by
suggesting there is finite mind space to cognitively work out more than
one language and a finite number of waking hours to use more than

one language, and hence where one language wins the other must lose.
Using the lens of the multi-competence perspective to critically exam-
ine these concerns and pitfalls would be beneficial for these other fields
as well. The SLA research community is best equipped to address the
challenges of measuring what looks like the incipient and very hetero-
geneous language competencies of L2 users who are simultaneously
seamless users of their stronger (longer used) languages, and thus to
build systematic accumulation of knowledge about same-participant
variegated and interacting skills across their languages. Therefore, the
full and critical pursuit of this domain of research by the field would
contribute in no uncertain way to make SLA knowledge more transdis-
ciplinarily relevant and more valuable to other fields (Ortega 2013).
3.5 Rejecting native speakers as normative comparison

benchmarks for interpretation
As part of a bilingual perspective into the late-timed learning of addi-

tional languages, multi-competence has explicitly and loudly rejected
the use of the native speaker as the benchmark for linguistic compe-
tence. Premise 2 stipulates that multi-competence does not depend on the
monolingual native speaker and specifically demands that the linguistic
competence of L2 users be studied in its own right. The overreliance on
the idealized construct of “the native speaker” as the ultimate arbiter of
what counts as complete linguistic knowledge and competent and successful
language use has received extensive critique in the realm of critical
applied linguistics (e.g. May 2014) but, as Seidlhofer (2001) argued, it is
still widespread among scholarly circles in the whole of applied linguis-
tics. This is despite vigorous efforts at addressing it through helpful
alternative understandings of phenomena, for example, lingua franca
or intelligibility. The construct of multi-competence itself was a psycho-
linguistic response to provide one such alternative understanding in
SLA for the traditional construct of linguistic competence or ability
(Cook 1991). Yet, comparisons of L2 learners against monolingual
natives of the target language have always been the unquestioned
staple of traditional SLA research (note its presence in all designs in
Table 3.1, except for Type 0 and Type 3).
Nobody would deny that L1 and L2 users are very different. Just how
different are they, indeed? Dinsmore (2006) meta-analyzed 16 SLA studies
investigating Universal Grammar issues published between 1981 and 1999
and was able to pinpoint a numerical average difference: monolingual and
bilingual adult users differed in their grammaticality intuitions when
judging syntactic phenomena in a given target language by over one
standard deviation unit on average (d = 1.25). Put differently, the L1 mono-
lingual user groups in those 16 studies performed on average at the 84th

62 LOURDES ORTEGA
percentile, if the L2 user groups are hypothetically placed at the 50th

percentile. However, what is disputed is what such a difference may tell
us about the nature of L2 bilingual versus L1 monolingual competence.
The readiest answer to this question in SLA to date has been that L2
learners fail to achieve complete linguistic success. This answer is only
possible because the monolingual L1 baselines are taken to be the norm
and benchmark for all forms of linguistic competence, including that of L2
users. Thus, the comparison is doomed to yield not just difference but also
disadvantage for the L2 learners. And this disadvantage is interpreted
as failure. And this failure is reified into a definitional feature of L2
competence.
To counter these deficit effects, the multi-competence perspective has
proposed two strategies. One has been to avoid the descriptor “L2 learner”
because it is a deficit-prone label and to adopt instead “L2 user” as a more
precise and less demeaning or patronizing term. The other has been to
reject monolingual native speaker comparisons as the sole or principal
basis for all research interpretations. Have these strategies been taken up
in the mainstream of the field?
My colleague Casey Keck and I set out to examine the impact of the first
suggestion by tracking the frequency of occurrence of the terms “L2
learner(s)” and “L2 user(s)” in the contemporary disciplinary discourse of
applied linguists (Keck and Ortega 2013). We did so in a corpus we created,
comprising over 6.5 million words and over 900 articles published
between 2005 and 2012 across 11 well-known international journals in
applied linguistics (including main SLA journals). We found a staggering
raw frequency difference of 20,312 “L2 learner(s)” instances versus 145
“L2 user(s)” tokens. “Learners” were invoked an average of 30 times in a
typical research article of about 10,000 words, whereas “L2 user” as a term
occurred a meager 0.2 times per article on average.
Obviously, the label “L2 user” has not caught on in traditional SLA. But
might we see some more impact of multi-competence in the proposed
rejection of monolingual native speaker comparisons? Are traditional
SLA researchers gradually awakening to at least this other piece in the
multi-competence critique against deficit construals of adult L2 users?
My analysis of the present situation is both optimistic and pessimistic.
As I see it, Premise 2 of multi-competence has made inroads into tradi-
tional SLA by succeeding in making theoretically diverse SLA researchers
open up to debating the necessity and felicity of this practice. On the
other hand, the debate is like a glacier, gradually formed over 50 years of
mounting awareness and still proceeding at a discouragingly slow pace.
Occasional breakthroughs are achieved (Bley-Vroman 1983; Cook 1991;
Klein 1998), as if icebergs broke off the glacier. But the debating of such
insights remains superficial because it limits itself (thus far at least) to
examining just the icebergs’ visible tips, ignoring what are very large
masses of submerged, hidden-to-the-eye ideologies.

3.6 Of glaciers and icebergs: Current debates of Premise 2

in traditional SLA
There has been a recent willingness among SLA researchers who work
outside the multi-competence perspective to at least discuss publicly the
issue of native speaker comparisons in SLA research, and this is a welcome
development. For example, in 2011, the question Is there a future for the
native speaker in SLA research? was examined at the Language Learning
Roundtable organized by Niclas Abrahamsson and Manne Bylund at the
EuroSLA conference hosted at the University of Stockholm; and a few
months later the issue of native speaker versus alternative comparisons
took up a good part of the presentations and audience discussion in an
invited colloquium on the more encompassing question (Why) does SLA
need a bilingual turn? organized by myself at the 30th Second Language
Research Forum hosted at Iowa State University. Two published reflections
by Slabakova (2013) and Birdsong and Gertken (2013) bear witness to the
field’s new openness in contending with the issue of appropriate baselines
and comparisons for SLA, which has always been the Trojan horse of
Premise 2 of multi-competence.
A prominent voice from a formal linguistic SLA perspective, Slabakova
(2013, pp. 53–54) offers a brief but explicit and forward-looking reflection.
She concedes that “[t]here are many reasons why monolingual–bilingual
comparisons look more and more like comparing apples and oranges”
(p. 53). In this she seems persuaded by the arguments found in neu-
roscience on cognitive advantages of bilingualism (e.g. Bialystok, Craik
and Luk 2012) and in the psycholinguistic research on cross-linguistic
interactions (e.g. Kroll, Bogulski and McClain 2012; van Hell and Tanner
2012). In her commentary, she adds a third argument emerging from
some sectors of generative SLA (Sorace 2011a also comes to mind), namely
that “the input of bilinguals and monolinguals is too varied for direct
comparisons to be justified” (p. 53). Another important voice in SLA is
that of Birdsong, who has been a long-term, even-tempered critic of the
practice of interpreting evidence on the critical period hypothesis by
direct comparison of native and non-native speaker performances. In a
recent article devoted to the topic of comparisons, Birdsong and Gertken
(2013) admonish that bilinguals are not monolinguals, and, citing
Slabakova, that furthermore the linguistic experience of monolinguals
and bilinguals is so different that native–non-native comparisons are
greatly compromised. Clearly siding with Premise 2 of multi-competence,
they also warn that proclamations that late bilinguals are deterministi-
cally doomed to failure just because they can or cannot achieve
monolingual-like competence in the L2, operationalized as performing
within the ranges of native speaking participants, rest on a gross mischar-
acterization of both bilingual and monolingual competence. Thus, on the

64 LOURDES ORTEGA
score as to whether the practice to use direct native–non-native speaker

comparisons has psycholinguistic validity, both Slabakova and Birdsong
and Gertken seem to agree with each other and with the multi-competence
perspective that it does not.
Ultimately, however, the real question that seems to create great
anxiety in this glacial (slow-moving yet unstoppable) debate in tradi-
tional SLA is whether the field might be able to find alternative (or even
just additional) baselines or yardsticks for their empirical programs. On
this score Birdsong and Gertken (2013) answer negatively that no suffi-
ciently good alternatives can be had at present, as much as this is “an
inconvenient truth” (p. 114) – because presumably they wish we had
some. For this and some other reasons I will not discuss here, they
extend their conditional support to the traditional reliance on native
speaker comparisons, provided the purposes to which they are put are
appropriately qualified in the particular studies. On the other hand,
Slabakova (2013) is willing to look hypothetically for alternative com-
parative yardsticks, and specifically offers two: (a) from-birth linguisti-
cally balanced bilinguals, and (b) extremely advanced or near-native late
bilinguals (p. 54). The first alternative has been advocated recently by
Sorace (2011b) and much earlier in the field by Singleton (2003), among
others. The second alternative is the one Slabakova herself favors. Her
final recommendation is to continue the conversation as to old and
new comparisons for interpretive benchmarking that may benefit the
field in the future. She forecasts the concurrent use of multiple yard-
sticks (i.e., monolingual, early bilingual, and late bilingual) is probably
ideal, as in her opinion this multiplicity of baselines is likely to afford
SLA researchers maximum interpretive power.
3.7 Tips of icebergs, and the difficulties of linguistic

definitions of bilingualism
As welcome as these critical reflections within SLA may be, it is clear that
some traditional premises remain intact in the ongoing debates and that
only the tips of enormous icebergs are seen and discussed. One, the faith
in the psycholinguistic reality of the native speaker seems unshakable,
contra the by now widespread sociolinguistic and poststructuralist posi-
tions that it is only an idealization. Two, the confidence in the research
utility of the monolingual native speaker as a yardstick (albeit now
proposed by some as just one of several yardsticks) remains unscathed.
This is seen in Birdsong and Gertken’s (2013) and Slabakova’s (2013)
positions, already discussed, as well as in an innovative empirical
exploration of the value of L1 baselines by Andringa (2014), who like
these other scholars works outside the multi-competence framework.
He has gone beyond most SLA research to date in his willingness to

question the validity of a uniform, unidimensional native speaker by

tackling the issue empirically.
And indeed with his empirical findings, Andringa (2014) has provided
robust evidence that education level creates different native speaker
benchmarks, and that variation within monolingual native speaker per-
formance (and not just within L2 users) can have important consequences
for the validity of the existing comparisons. For example, in his tests of
Dutch vocabulary processing, 19% (22 of 114) of L2 users who scored below
the range of 34 highly educated L1 users did make it within the native
range when the comparison was against a more varied sample of 77
L1 users representing the full educational spectrum typical of the
Netherlands. That is, in many studies, these 22 L2 users would have been
classified as non-nativelike by comparison to the highly educated sample
but as nativelike by comparison to the representative-education L1
sample. He concluded that native speaker performance must be (a) empiri-
cally studied better, (b) sampled more robustly in terms of n-size, and
(c) selected to be more representative of the intended speech community.
To these three recommendations, a fourth can be added from Brown
and Gullberg (2012). These researchers noted that many so-called native
speakers in the L1 groups recruited in SLA studies do in fact have knowl-
edge of more than one language, however minimal and dormant. They
therefore recommend collecting full demographic information from not
only L2 participants but also L1 participants. They encourage all SLA
researchers to think of both types of participants as “maximally” or
“functionally” monolingual or bi/multilingual, on a continuum that
must be empirically demonstrated.
All these recommendations are well worth heeding, if research bench-
marks are to be extracted from monolingual native speaker groups. But
the crux of the matter remains whether one needs to question wholesale
the validity of the native speaker as a construct and as a norm-setting
heuristic.
To my mind, the biggest tip-of-the-iceberg problem is that, even when
the validity of nativeness and its norm-setting uses is questioned, another
traditional premise that remains unhelpfully intact is privileging of
the linguistic definition of bilingualism as “native-like control of two
languages” (Bloomfield 1933, p. 56), which takes precedence over other
definitions that also enjoy standing and currency in this other field (see
the many definitions discussed by Li Wei in his introduction to the 2007
paperback edition of the Bilingualism Reader).
A variety of non-linguistic definitions of bilingualism are probably at
the center of relevance for understanding the competencies of (probably
all) bi/multilinguals, including many heritage language learners with
heterogeneous experiences and profiles (Rothman and Treffers-Daller
2014), but also approximately one-fourth of from-birth bilinguals who
will speak, as opposed to comprehend, only one of their languages (de

66 LOURDES ORTEGA
Houwer 2015). The gray areas a linguistic definition of bilingualism

creates for research are formidable. For example, the seminal study by
Peal and Lambert (1962) is often cited as having countered the stigma of
deficit that had been associated with bilingualism up until the 1960s and
as being the initiator of the important line of research that has by now
catapulted the intellectual and cognitive advantages of bilinguals into
the public imagination. Incidentally, but perhaps not coincidentally, this
study adopted a Type 5 design (see Table 3.1). What is less often cited
or noted is that of the initial pool of 364 carefully controlled and tested
ten-year-old children from six Montreal French schools, 200 were
discarded from the final study. The reason? They were not “equally skilled
in French and English” (p. 8), as seen in their dissimilar scores in French
and English across the exhaustive battery administered by the researchers.
In other words, because these 200 children may have been dominant in
one of their two languages, rather than linguistically balanced in both,
Peal and Lambert felt they “could not be unambiguously classified as
either monolingual or bilingual” (p. 8) and had to be discarded.
As the Peal and Lambert (1962) illustration clearly reveals, linguistic
definitions of what it means to be bilingual are in essence shaped by
complex and contradictory biases; more likely than not, they are informed
by expectations that “balanced” linguistic bilingualism will look like the
sum of two monolingual competences, particularly given that judgments
about such balance have to date most often been made on the basis of
monolingual baselines. Since L2 users, in particular, are not easily accom-
modated in monolingually shaped linguistic definitions of bilingualism,
and since many are indeed dominant in one of their languages (often their
L1 but sometimes their L2: Birdsong 2014), little progress will be made in
advancing Premise 2 of the multi-competence perspective until the field
of SLA as a whole is willing to grapple with non-linguistic definitions of
bi/multilingualism as relevant for their studies and for the goals that drive
SLA research.
Are there foreseeable alternatives to monolingual native baselines for
the future, given the bleak picture I have painted of the current debate? It
will undoubtedly be eye-opening to compare adult-emergent bilinguals
to from-birth developing bilinguals, as Sorace (2011b) has proposed, and
it will also be informative (and liberating) to employ samples of very
advanced L2 users as comparative yardsticks, as suggested by Slabakova
(2013). Nevertheless, some caveats must be kept in mind when implement-
ing either or both proposals. First, from-birth exposure does not guarantee
“bilingual competence” in a linguistic definition of the term (de Houwer
2015) and thus carefully planned sampling strategies and well-reported
controls and sensitivity analyses of proficiency, dominance, and patterns
of language use will be called for. This will place considerable additional
demands on SLA researchers. Second, the problem of the prevalent deficit
orientation will not go away if once again we elevate nativeness to arbiter

of all other forms of language competence, only now bilingual nativeness.

Third, nothing much may change if, behind the definition of “very
advanced” or “near-native,” once again we find monolingual native
benchmarks, for example, as adopted in the literature on L2 phonology
and pronunciation with the construct of intelligibility (e.g. Derwing and
Munro 2013), in the literature on the advanced learner reviewed by Byrnes
(2012), and in the recent definition of near-native speakers as speakers
characterized by unperceivable (but scrutinizable) non-nativeness
(Abrahamsson and Hyltenstam 2009). Leniency towards a different type
of competence does not equate to egalitarian treatment of that difference.
The headways and limitations of the baselines debate notwithstanding,
I would argue that any impact of Premise 2 on traditional SLA will be
superficial unless wider issues pertaining to the ultimate goals of the
field are recognized and unless ideologies of language that harm the field
are openly discussed (Ortega 2013, 2014).
3.8 Subordinating comparisons and their ideological roots
The overreliance on the idealized construct of the “native speaker” as

the ultimate arbiter of what counts as complete linguistic knowledge and
competent and successful language use is supported by nativespeakerism
(Holliday 2006), an ideological belief in the linguistic superiority of native
speakers held by researchers as much as laypeople. It has received exten-
sive analysis and critique in the realm of critical applied linguistics (e.g.
May 2014). Cook also denounces the comparison as discriminatory and
prejudiced when he says:
The nub of Labov’s sociolinguistic argument about discrimination is that
one group should not be measured against the norm of another group
(Labov, 1969) . . . Labov’s argument is that people who speak differently
from some arbitrary group are not using grammars that are better or
worse: they are just speaking differently . . . All branches of linguistics
accept the creed of “difference not deficit” with rare exceptions.
Why then should L2 users alone be singled out as deficient for being
what they are, bilinguals, and not what they are not, monolinguals?
(Cook 1997, p. 43)
Indeed, I would like to argue that the problem has ideological roots
and is not a problem of comparisons, but of subordinating comparisons.
I use the term subordinating comparisons to refer to comparisons made
under explicit or implicit assumptions about a superiority–inferiority
relation in the elements compared, where subordination can be only
ideologically but not logically justified. A subordinating comparison
may be better than no comparison in certain contexts and historically
it may serve to advance knowledge, but it is bound to lead to distorting
knowledge that eventually needs revision.

68 LOURDES ORTEGA
A good illustration of what subordinating comparisons are and what

they do to research knowledge can be found in the case of articles as a
part of speech (as recounted by Thomas 2004, p. 57). In the thirteenth
century, the speculative grammarians at the University of Paris were
faced with the task of resolving the mysterious discrepancy between
Latin, which is a no-article language, and Greek, which has an article
system. After much study, they concluded (a) there are eight universal
parts of speech, which excludes articles, and (b) Greek articles are case
markers, which are needed because (c) Greek nouns are indeclinable. In
hindsight, contemporary linguists know these thirteenth-century gram-
marians erred in all three counts. These errors are explained as limita-
tions in what they could envision within the bounds of their best
knowledge: they lived in a time when Latin was still considered a
superior language, and thus the necessary and only possible bench-
mark against which to understand all other languages. The adherence
in current traditional SLA knowledge to native speakers as the natural,
necessary, and only possible benchmark against which to understand the
competence of L2 users is an analogous case of a hypothesis space
limited severely by a subordinating comparison mindset.
Subordinating comparisons are difficult to unmask. Rothman and
Iverson (2010, pp. 42–44) provide an uncanny case in point pertaining to
the subjunctive. It has been empirically observed that in US English the
subjunctive is used more frequently in contexts where in UK English
modals (should, shall, may) are preferred. (For example, where a Yankee
may say It is important that he be told, a Brit may say It is important that he
should be told.) What would the enlightened linguist want to conclude from
this observed empirical difference? Would we want to say that US English
over-uses the subjunctive or that it under-uses modals? Now, it has also
been empirically observed that heritage (bilingual) speakers of Spanish
use the subjunctive in certain functional contexts less often than L1
(monolingual) speakers of Spanish do. Why are in this case SLA researchers
so willing to conclude out of this comparative finding that the heritage
grammars exhibit “incompleteness”? Several ideological processes nour-
ish the unashamed, unexamined practice of subordinating comparisons in
SLA: nativespeakerism, language purity, linguicism and monolingualism.
These ideological processes allow researchers to assume as a natural
(scientific!) truth that L1 users hold a superior kind of linguistic compe-
tence which is more perfect, more authentic, and more legitimate than
that of L2 users, just as the thirteenth-century speculative grammarians
assumed that Latin provided a better test case for cracking the inner
workings of grammar than any other language could. It is not that Latin
is the wrong choice for a subordinating comparison in the context of
explaining all grammars. In fact, any other language that could be desig-
nated as a superior frame of reference would have led the speculative
grammarians astray, if not in the issue of articles, in some other issue.

More to the point, eventually the comparative method must be outgrown

in any discipline, and other methods must be envisioned and added to
enable inquiry into phenomena in their own right. Cross-linguistic
typology is a vibrant branch of contemporary linguistics that employs
(non-subordinating) comparison as its main method. But contemporary
linguistics in many other subfields resorts to many other methods that
do not entail comparisons and baselines. So must SLA researchers even-
tually learn to study late bilingualism in its own right, as Premise 2 of
multi-competence demands.
I remain convinced that the beneficial impact of multi-competence as
a perspective will not be fully realized until the mainstream of the field
of SLA is willing to confront the ideological nature of the processes that
support subordinating native–non-native comparisons and lead to dis-
tortions of knowledge. Consider how the most oft-asked, broadest ques-
tion in SLA, Why is learning a language in adulthood so different from, and so
much more difficult than, learning the first language? juxtaposes adults and
children and L2s and L1s in which appears to be a logical and balanced
comparison. But the question quickly loses meaning if one translates it
under the perspective of multi-competence as Why is becoming bilingual
later in life so different from, and so much more difficult than, acquiring one
language only during the first few years of life? Now the question has turned
into an impossible comparison, because nothing is held constant: the
timing of learning varies (later in life versus from birth or very early) and
the number of languages involved in the learning and its outcomes
varies (more than one language versus only one). Thus, there is nothing
left for a rational comparison. Likewise, what is perhaps the most
popular questions in SLA as well as in cognitive science and the popular
imagination, Is it possible to achieve native-like competence in a second lan-
guage?, loses its appeal if one translates it as: Is it possible to achieve
monolingual-like competence in more than one language? It suddenly repre-
sents an oxymoron, since there is little point in understanding why
bilinguals are not monolinguals, once one takes seriously the compel-
ling evidence from bilingual studies (e.g., the four-volume anthology in
Li Wei 2009) that they are not. Most SLA researchers have been insulated
from recasting these research questions in the new ways that multi-
competence encourages because of the deeply held belief that monolin-
gual competence is of a superior, purer kind, linguistically speaking,
which subsequently allows the field to construe L2 users “as aspiring
monolinguals of the new language” (Ortega 2014, p. 36).
It is only if the work of ideologies is confronted that these beliefs and
practices in SLA come to be seen as having unacceptable consequences for
the validity and the ethics of the field. From a validity standpoint, the
knowledge that has accumulated over now almost five decades becomes
suspect, because it was generated by applying the mirage of the native
speaker ideal as the golden standard against which to both make sense of

70 LOURDES ORTEGA
L2 development and proclaim success or failure for late bilinguals. From

an ethical standpoint, such research becomes bankrupt, because it con-
structs L2 users as deficient and doomed to failure; moreover, these are
persuasive constructions not only for the research community, who end
up believing in them as natural facts, but also for educators and for L2
users themselves, who may internalize the deficit and enact it in class-
rooms and in their lives.
3.9 Conclusion
The disregard for the total language knowledge of L2 users together

with the isolation of knowledge of L2 as the only relevant object
of inquiry (both against Premise 1 of multi-competence), and the
entrenchment of the native–non-native speaker subordinating compar-
ison together with the construction of late bi/multilinguals as linguis-
tically deficient (both against Premise 2 of multi-competence) have
guided the research questions asked in SLA research and indeed shaped
the whole raison d’être of the field. The perspective of multi-
competence has helped make sense of these problems and has offered
alternatives for the future. While the inroads of multi-competence into
traditional SLA are visible and by now indelible, the time may have
come for the whole field to rethink both our research questions and
our disciplinary goals.
SLA as a field may now be ready to take the next two logical
empirical steps. First, the field needs to take stock of the conse-
quences of studying the L2 in isolation from the other languages of
bi/multilingual minds and of interpreting all findings by reference to
monolingual native yardsticks. This can be done empirically with new
research programs as well as by systematically synthesizing the impact
of these practices on findings and interpretations across domains. We
must do both in order to ascertain precisely where and how (as some
of us believe) the status quo might have led to validity problems
where accumulated knowledge has been hidden or distorted. And
second, targeted research programs need to be initiated that explore
alternatives to native speaker baselines as an investment in the future
of the field. Eventually, the search for better baselines will be moot
and the reliance on the comparative method will be outgrown, when
SLA researchers become able to envision methods for the study of
late-timed adult bilingualism in its own right, as the multi-
competence perspective demands. I remain convinced, however, that
the beneficial impact of multi-competence will not be fully realized
until the mainstream of the field of SLA is willing to confront the
ideological nature of the processes involved in prevailing research
practices and theories.

Note
1. Of course, there are many variations around the designs listed in

Table 3.1. An ingenious variation of Type 5 design is offered by
Gathercole and Moawad (2010). These researchers collected data from
two monolingual baseline groups (Arabic and English) in their single
language. They also collected single-language data from their L2 user
participants, who were early versus late (i.e. pre- versus post-pubescent
starting) EFL college students of English literature and translation in
Saudi Arabia. However, half of the L2 users provided L1 data (Arabic)
and half L2 data (English). In this way, a between-subjects design was
followed, but one that in essence produced evidence by (different
groups of) L2 users on all the L2 user participants’ languages, in a single
language per individual. This Type 5 modification requires a large
number of L2 participants (in order to obtain sufficient subjects per
cell and robust evidence on their two languages), but it has two advan-
tages: (a) it makes the data collection sessions shorter for each indivi-
dual participant and (b) it preempts the problems of counterbalancing
of language order and addressing boredom or practice effects when
bilingual/L2 user participants are asked to do the same tasks twice,
one time in each language.
References
Abrahamsson, N. and Hyltenstam, K. 2009. ‘Age of L2 acquisition and

degree of nativelikeness: listener perception versus linguistic scrutiny’,
Language Learning 59, 249–306.
Andringa, S. 2014. ‘The use of native speaker norms in critical period
hypothesis research’, Studies in Second Language Acquisition 36 (3), 565–596.
Bialystok, E., Craik, F. I. M. and Luk, G. 2012. ‘Bilingualism: consequences
for mind and brain’, Trends in Cognitive Sciences 16(4), 240–250.
Birdsong, D. 2014. ‘Dominance and age in bilingualism’, Applied Linguistics.
doi:10.1093/applin/amu031
Birdsong, D. and Gertken, L. M. 2013. ‘In faint praise of folly: a critical
review of native/non-native speaker comparisons, with examples from
native and bilingual processing of French complex syntax’, Language,
Interaction and Acquisition 4, 107–133.
Bley-Vroman, R. 1983. ‘The comparative fallacy in interlanguage studies:
the case of systematicity’, Language Learning 33, 1–17.
Brown, A. and Gullberg, M. 2008. ‘Bidirectional crosslinguistic influence in
L1–L2 encoding of manner in speech and gesture’, Studies in Second
Language Acquisition 30, 225–251.

72 LOURDES ORTEGA

variation in clausal packaging in Japanese’, Second Language Research 28,
415–442.
Brown, A. and Gullberg, M. 2013. ‘L1–L2 convergence in clausal pack-
aging in Japanese and English’, Bilingualism: Language and Cognition 16,
477–494.
Bylund, E. and Athanasopoulos, P. 2014. ‘Linguistic relativity in SLA:
towards a new research program’, Language Learning 64 (4), 952–985.
Byrnes, H. 2012. ‘Advanced language proficiency’, in S. M. Gass and
A. Mackey (eds), Handbook of Second Language Acquisition, pp. 506–520.
New York: Routledge.
Cenoz, J., Hufeisen, B. and Jessner, U. 2001. Cross-linguistic Influence in
Third Language Acquisition: Psycholinguistic Perspectives. Clevedon, UK:
Cook, V. J. 1975. ‘Strategies in the comprehension of relative clauses’,
Language and Speech 18, 204–212.
Cook, V. J. 1993. ‘Wholistic multi-competence: jeu d’esprit or paradigm
shift?’, in B. Kettemann (ed.), Current Issues in European Second Language
Acquisition Research, pp. 2–8. Tübingen, Germany: Gunter Narr Verlag.
Cook, V. J. (ed.) 2003. Effects of the Second Language on the First. Clevedon, UK:
Cook, V. J. 2012. ‘Multi-competence and the native speaker’, in
C. A. Chapelle (ed.), The Encyclopedia of Applied Linguistics, pp. 3768–3774.
Malden, MA: Blackwell.
Core, C., Hoff, E., Rumiche, R. and Señor, M. 2013. ‘Total and conceptual
vocabulary in Spanish–English bilinguals from 22 to 30 months: impli-
cations for assessment’, Journal of Speech, Language, and Hearing Research
56, 1637–1649.
De Angelis, G. and Dewaele, J.-M. (eds) 2011. New Trends in Crosslinguistic
Influence and Multilingualism Research. Clevedon, UK: Multilingual Matters.
De Bot, K., Lowie, W., Thorne, S. L. and Verspoor, M. 2013. ‘Dynamic
Systems Theory as a comprehensive theory of second language develop-
ment’, in M. del Pilar Garcı́a Mayo, M. J. Gutierrez Mangado and
M. Martı́nez Adrián (eds), Contemporary Approaches to Second Language
Acquisition, pp. 199–220. Amsterdam: John Benjamins.
De Houwer, A. 2009. Bilingual First Language Acquisition. Bristol, UK:
De Houwer, A. 2015. ‘Harmonious bilingual development: young families’
well-being in language contact situations’, International Journal of
Bilingualism 19, 169–184.

De Leeuw, E., Opitz, C. and Lubińska, D. (eds) 2013. The Dynamics of First
Language Attrition across the Lifespan, special issue of International Journal of
Bilingualism 17 (6).
De Zeeuw, M., Verhoeven, L. and Schreuder, R. 2012. ‘Morphological
family size effects in young first and second language learners: evidence
of cross-language semantic activation in visual word recognition’,
Derwing, T. M. and Munro, M. 2013. ‘The development of L2 oral language
skills in two L1 groups: a seven-year study’, Language Learning 63,
163–185.
Dinsmore, T. H. 2006. ‘Principles, parameters, and SLA: a retrospective
meta-analytic investigation into adult L2 learners’ access to Universal
Grammar’, in J. M. Norris and L. Ortega (eds), Synthesizing Research on
Language Learning and Teaching, pp. 53–90. Amsterdam: John Benjamins.
Five Graces Group: Beckner, C., Blythe, R., Bybee, J., Christiansen, M. H.,
Croft, W., Ellis, N. C., . . . Schoenemann, T. 2009. ‘Language is a complex
adaptive system: position paper’, Language Learning 59 (supplement 1),
1–26.
Flege, J. E. 1987. ‘The production of “new” and “similar” phones in a
foreign language: evidence for the effect of equivalence classification’,
Journal of Phonetics 15, 47–65.
Garcia, O. and Li Wei 2014. Translanguaging: Language, Bilingualism and
Education. Basingstoke, UK: Palgrave Macmillan.
Gass, S. M. and Selinker, L. (eds) 1983. Language Transfer in Language Learning.
Rowley, MA: Newbury House.
Gathercole, V. C. M. and Moawad, R. A. 2010. ‘Semantic interaction in early
and late bilinguals: all words are not created equally’, Bilingualism:
Language and Cognition 13, 385–408.
Grosjean, F. 1989. ‘Neurolinguists, beware! The bilingual is not two mono-
linguals in one person’, Brain and Language 36, 3–15.
Grosjean, F. 2008. Studying Bilinguals. Oxford University Press.
Grüter, T., Lieberman, M. and Gualmini, A. 2010. ‘Acquiring the scope of
disjunction and negation in L2: a bidirectional study of learners of
Japanese and English’, Language Acquisition 17 (3), 127–154.
Han, Z.-H. and Cadierno, T. (eds) 2010. Linguistic Relativity in SLA: Thinking for
Speaking. Clevedon, UK: Multilingual Matters.
Hoff, E. 2003. ‘The specificity of environmental influence: socioeconomic
status affects early vocabulary development via maternal speech’, Child
Development 74, 1368–1878.
Holliday, A. 2006. ‘Native-speakerism’, ELT Journal 60, 385–387.
Jarvis, S. 2000. ‘Methodological rigor in the study of transfer: identifying L1
influence in the interlanguage lexicon’, Language Learning 50, 245–309.
Cognition. Mahwah, NJ: Lawrence Erlbaum.

74 LOURDES ORTEGA
Keck, C. and Ortega, L. 2013. Deficit views of language learners in applied

linguistics discourse: a corpus-based critical discourse analysis. Paper
presented at the 32nd Second Language Research Forum, October
31–November 2, Provo, UT.
Kellerman, E. and Sharwood Smith, M. (eds) 1986. Crosslinguistic Influence in
Second Language Acquisition. New York: Pergamon Press.
Klein, W. 1998. ‘The contribution of second language acquisition
research’, Language Learning 48, 527–550.
Kroll, J. F., Bogulski, C. A. and McClain, R. 2012. ‘Psycholinguistic perspec-
tives on second language learning and bilingualism: the course and
consequence of cross-language competition’, Linguistic Approaches to
Labov, W. 1969. ‘The logic of nonstandard English’, Georgetown Monographs
on Language and Linguistics 22 (1–22), 26–31.
Langacker, R. 2008. Cognitive Grammar: A Basic Introduction. New York:
Oxford University Press.
Li Wei (ed.) 2007. The Bilingualism Reader, second edition. London:
Routledge.
Li Wei (ed.) 2009. Bilingualism and Multilingualism: Critical Concepts in
Linguistics, four-volume anthology. London: Routledge.
MacSwan, J. 2000. ‘The Threshold Hypothesis, semilingualism, and other
contributions to a deficit view of linguistic minorities’, Hispanic Journal of
Behavioral Sciences 22, 3–45.
May, S. (ed.) 2014. The Multilingual Turn: Implications for SLA, TESOL and
Bilingual Education. New York: Routledge.
Odlin, T. 1989. Language Transfer: Cross-linguistic Influence in Language Learning.
New York: Cambridge University Press.
Ortega, L. 2009. Understanding Second Language Acquisition. London:
Routledge.
Ortega, L. 2013. ‘SLA for the 21st century: disciplinary progress, transdisci-
plinary relevance, and the bi/multilingual turn’, Language Learning 63
(S1), 1–24.
Ortega, L. 2014. ‘Ways forward for a bi/multilingual turn in SLA’, in S. May
(ed.), The Multilingual Turn: Implications for SLA, TESOL and Bilingual
Education, pp. 32–53. New York: Routledge.
Pavlenko, A. and Jarvis, S. 2002. ‘Bidirectional transfer’, Applied Linguistics
23, 190–214.
Peal, E. and Lambert, W. E. 1962. ‘The relation of bilingualism to intelli-
gence’, Psychological Monographs: General and Applied 76 (27), 1–23.
Pearson, B. Z., Fernández, S. C. and Oller, D. K. 1993. ‘Lexical development
in bilingual infants and toddlers: comparison to monolingual norms’,
Pulvermüller, F. 2013. ‘How neurons make meaning: brain mechanisms
for embodied and abstract-symbolic semantics’, Trends in Cognitive
Sciences 17, 458–470.

Ringbom, H. 1987. The Role of the First Language in Foreign Language Learning.
Clevedon, UK: Multilingual Matters.
Ringbom, H. 2007. Cross-linguistic Similarity in Foreign Language Learning.
Rothman, J. and Iverson, I. 2010. ‘Independent normative assessments
for bi/multilingualism, where art thou?’, in M. Cruz-Ferreira (ed.),
Multilingual Norms: Language Norming in Multilingual Contexts, pp. 33–51.
Frankfurt am Main: Peter Lang.
Rothman, J. and Treffers-Daller, J. 2014. ‘A prolegomenon to the construct
of the native speaker: heritage speaker bilinguals are natives too!’
Applied Linguistics 35, 93–98.
Rowe, M. L., Raudenbush, S. W. and Goldin-Meadow, S. 2012. ‘The pace of
vocabulary growth helps predict later vocabulary skill’, Child Development
83, 508–525.
Schachter, J. 1974. ‘An error in error analysis’, Language Learning 24,
205–214.
Schmitt, N. 2014. ‘Size and depth of vocabulary knowledge: what the
research shows’, Language Learning 65, 913–951.
Seidlhofer, B. 2001. ‘Closing a conceptual gap: the case for a description of
English as a lingua franca’, International Journal of Applied Linguistics 1,
133–158.
Silva-Corvalán, C. and Treffers-Daller, J. (eds), in press. Language Dominance
in Bilinguals: Issues of Measurement and Operationalization. Cambridge
University Press.
Singleton, D. 2003. ‘Critical period or general age factor(s)?’, in M. del Pilar
Garcı́a Mayo and M. L. Garcı́a Lecumberri (eds), Age and the Acquisition
of English as a Foreign Language, pp. 3–22. Clevedon, UK: Multilingual
Matters.
Slabakova, R. 2013. ‘Adult second language acquisition: a selective over-
view with a focus on the learner linguistic system’, Linguistic Approaches to
Sorace, A. 2011a. ‘Pinning down the concept of “interface” in bilingual-
ism’, Linguistic Approaches to Bilingualism 1 (1), 1–33.
Sorace, A. 2011b. The bilingual native speaker. Paper presented at the
Language Learning Roundtable on Is there a future for the native speaker in
SLA research? at the 21st Annual Conference of the European Second
Language Association (N. Abrahamsson and M. Bylund, conveners),
Stockholm.
Taylor, D. M. 2011. ‘Where it all began: a tribute to Wallace E. Lambert’,
Journal of Language and Social Psychology 30, 259–263.
Thomas, M. 2004. Universal Grammar in Second Language Acquisition: A History.
Treffers-Daller, J. 2011. ‘Operationalizing and measuring language domi-
nance’, International Journal of Bilingualism 15, 147–163.

76 LOURDES ORTEGA
Van Hell, J. G. and Tanner, D. 2012. ‘Second language proficiency and

cross-language lexical activation’, Language Learning 62, 148–171.
White, L. 1985. ‘The “pro-drop” parameter in adult second language
acquisition’, Language Learning 35, 47–62.
Yu, L. and Odlin, T. (eds) 2015. New Developments in Transfer Research. Bristol,

4
Not through a glass
darkly: refocusing the
psycholinguistic study
of bilingualism through
a “bivocal” lens
Jyotsna Vaid and Renata Meuter
[T]he abstraction [by Chomsky] to a monolingual speaker is of a

different kind from assuming that all speakers in a community
are effectively the same; it is simplifying the average mind
rather than finding an average mind, so to speak. If most
human beings are in fact multi-competent, monocompetence
is a misleading representation of the human species rather
than a convenient idealization. (Cook 1991, p. 114)
4.1 Introduction
In his preface to the 1992 edited volume Cognitive Processing in Bilinguals,

Richard Harris noted that “bilingualism is barely mentioned in most
cognitive psychology texts,” and that, for the most part (Paivio and Begg
1981; Taylor and Taylor 1990), even psycholinguistic texts “treat bilingualism
only very lightly, if at all” (Harris 1992, p. 9). This observation still stands.
With few exceptions (e.g. Paivio and Begg 1981; Taylor and Taylor 1990;
Fernandez and Cairns 2010), psycholinguistic textbooks give scant attention
to the topic; for example, in Introducing Psycholinguistics, the word “bilingual-
ism” is not given its own entry and there is only a single reference to “second
Preparation of this manuscript was facilitated by a Faculty Development Leave awarded to the first author by Texas A&M
University. We are grateful to Diana Ramirez, librarian at Texas A&M University, for guidance in data search, Caitlin Ohman
for initial assistance in data coding, and Jennifer Lange (Texas A&M University), Sara Berndt, and Ilse Kaiplinger
(Queensland University of Technology) for additional data searching and manuscript preparation assistance.

78 J Y OT S N A V A I D A N D R E N ATA M E U T E R
language processing” in the book’s index (Warren 2013). Where bilingualism

is brought up in psychology of language textbooks, it typically figures either
as a subtopic in language acquisition (e.g. Berko Gleason and Bernstein
Ratner 2012), or in the context of a discussion of the relationship between
language and thought (e.g. Carroll 2008).
The somewhat neglected status of bilingualism within cognitive psy-
chology may in part reflect a prevailing belief among psychologists that
human cognitive abilities in healthy adults are universal and unvarying.
The primary contexts in which individual differences in cognition have
been studied involve comparisons of healthy individuals with clinically
impaired ones or differences related to aging. The relative lack of interest
in bilingualism (until recently) may also, and perhaps largely, reflect
an implicit monolingual focus of psycholinguistics from the time of its
founding. This focus may have been bolstered by an unconscious bias on
the part of early researchers (many of whom were themselves monolin-
gual) towards a language model that fits their own reality and their nega-
tive perceptions of bilinguality (Pavlenko 2002).
Whatever the reasons, the continued persistence of a monolingual-as-
norm framework within cognitive psychology and psycholinguistics is all
the more curious given the inescapable fact that the majority of language
users are bilingual or multilingual rather than single language users.
Whereas it is customarily assumed that bi/multilingualism is a relatively
recent phenomenon, resulting from colonialism, globalization, and large
scale migration, recent scholarship shows that monolingualism is, in fact,
the more recent phenomenon (Yildiz 2012) and that bilingualism/multilingu-
alism was arguably the primal human condition in early small-scale speech
communities (Evans 2013). The fact that monolingualism has become the
dominant framework within which to view the world has resulted in a
tension in how bilingualism is to be treated, a tension that is becoming
more evident in the contemporary period, reflecting what Yildiz (2012)
calls “the post-monolingual condition.” Here we reflect on how this tension
is manifest in scholarship on bilingualism within the field of psychology.
Prior to the 1960s the study of language – in monolinguals and bilinguals
alike – had been examined from a neo-associationist learning perspective
(e.g. Ervin and Osgood 1954; Osgood and Sebeok 1954; Dil 1972; Lambert
1977). Psycholinguistics as a recognized subfield of psychology emerged in
the 1960s out of the so-called cognitive revolution in psychology, which
challenged the established behaviorist paradigm. The critique of the beha-
viorist approach to language, as formulated by Noam Chomsky in his
review of B. F. Skinner’s Verbal Behavior (Chomsky 1959), as well as a
growing acknowledgment by some neobehaviorists of the need to invoke
unobservable structures or processes (goals, plans, and mental representa-
tions) in making sense of behavior, led to a new way of conceptualizing
language and cognition more broadly (Gardner 1985). In this new concep-
tualization, language was cast in mentalist terms, as a shared symbolic

Not through a glass darkly 79
code used by an ideal speaker/hearer. Knowing a language came to mean

having a tacit linguistic competence about rules specifying how constitu-
ents could be combined to generate permissible, well-formed utterances in
the language (Carroll 2008).
Importantly, the “ideal speaker/hearer” was theorized as a native
speaker of the language and one who was in effect monolingual.
Psycholinguistics – the empirical study of how language is acquired,
understood, produced, represented, and/or impaired in the mind of an
individual language user – was thus largely rooted in a monolingual con-
ception of the canonical language user (Grosjean 1985). A by-product of
this view was that bilinguals were cast as atypical language users and the
study of bilingualism per se was thus not considered of mainstream interest
among researchers in North America. This was particularly the case in the
United States; the study of bilingualism had a somewhat different trajec-
tory in Canada, for other reasons, given that country’s different history
and policies.
Despite its relative neglect within mainstream US psychology and psy-
cholinguistics, social psychological and educational research on bilingual-
ism has in fact had a long history dating back to the early twentieth
century (Grosjean 1982; Hakuta 1986). Quite independently of this work,
there is also an early – largely European – body of work on aphasia in
bilinguals or polyglots which dates back even further, to the late 1880s
(Albert and Obler 1978; Paradis 1980).
Several influential empirical investigations of cognitive, social, and
educational aspects of bilingualism were carried out from the late 1950s
to the 1990s by the McGill University social psychologist Wallace E.
Lambert (Dil 1972; Lambert 1977; Vaid, Paivio, Gardner and Genesee
2010). Influenced in turn by theoretical accounts of the bilingual’s linguis-
tic systems articulated by Uriel Weinreich (1953), and by Susan Ervin-Tripp
and Charles Osgood (Ervin and Osgood 1954), Lambert’s pioneering
research on bilingualism, including studies of aphasia in bilinguals, the
measurement of language dominance, bilingual linguistic memory, bilin-
gual information processing, and language lateralization in bilinguals,
launched further work on these and related topics (see Keatley 1992 for a
history of cognitive research in bilingualism). By the late 1980s several
empirical studies of bilingualism appeared in psychology journals and a
number of volumes gradually emerged (Grosjean 1982; Vaid 1986/2014;
Reynolds 1991; Harris 1992; de Groot and Kroll 1997).
It was during this period that the notion of a holistic view of bilingual-
ism was first proposed by the psycholinguist François Grosjean (1985,
1989). Grosjean (1989) strongly cautioned against what he saw as a ten-
dency among some researchers to view bilinguals as merely “the sum of
two monolinguals” (p. 3) and against using monolinguals as the explicit or
implied standard against which bilinguals’ performance was to be evalu-
ated. Grosjean (1985) presented a case for the value of studying bilinguals

as specific speaker–hearers rather than as deviant language users. A holis-

tic (as opposed to a fractional) perspective, for Grosjean (1985), would be
one in which – among other things – bilingual-specific phenomena such as
code-switching (among bilingual interlocutors) would be foregrounded
and emphasis would be placed on how the two languages interact in the
bilingual mind.
At around the same time that Grosjean argued for a holistic view of
bilinguals, within the subfield of applied linguistics that focused on sec-
ond language acquisition, Vivian Cook (1991, 1992) advanced the notion of
“multi-competence.” This term was proposed to characterize the mental
functioning of multilingual speakers without having to make reference to
a native speaker standard. Cook argued that bilinguals (and multilinguals
more generally) are multi-competent users of language rather than imper-
fect second language learners who are not quite or not yet monolingual.
The multi-competence view of second language acquisition and, by exten-
sion, bilingualism has had a wide ranging impact not only in applied
linguistics (see Kecskes 2010) but also in allied fields such as sociolinguis-
tics, multilingualism (Franceschini 2011), language teaching, and
education policy.
In the 25 or so years since Cook’s and Grosjean’s critical essays first
appeared, there has been an explosion of research on linguistic, cogni-
tive, and neurocognitive aspects of bilingualism and multilingualism.
Given this steadily increasing research interest in psycholinguistic and
neurocognitive aspects of bilingualism, one could say that inquiry into
bilingualism has entered the mainstream. The mainstreaming of bilin-
gualism as a focus of inquiry is evident in the proliferation of confer-
ences and symposia devoted to it, in the inclusion of bilingualism as one
of the recognized topics at annual conferences in the psychology of
language (e.g. Architectures and Mechanisms in Language Processing;
the Society for the Neurobiology of Language) and in experimental psy-
chology more broadly (e.g. the Society for Psychonomic Science), and by
the growing representation of bilingualism within mainstream journals
in psycholinguistics (e.g. Journal of Memory and Language) and/or cognitive
psychology (e.g. Cognition). Finally, it is indicated by the number of hand-
books and edited volumes on psycholinguistic, cognitive, and/or cogni-
tive neuroscience treatments of bilingualism (Wei 2000; Altarriba and
Heredia 2008; Kroll and de Groot 2009; Cook and Bassetti 2011; de Groot
2011; Grosjean and Li 2013; Hernandez 2013; Pavlenko 2014), specialized
journal outlets such as Bilingualism: Language and Cognition, the
International Journal of Bilingualism, and the International Journal of
Multilingualism, as well as the emergence of research centers on bilin-
gualism situated in bilingual communities (e.g. The Centre for Research
on Bilingualism, in Bangor, Wales, and the Basque Center on Cognition,
Brain and Language in San Sebastián). The aim of research at the Basque
Center, as stated on its website, is “to unravel the neurocognitive

mechanisms involved in the acquisition, comprehension, and produc-

tion of language with special emphasis on bilingualism and multilin-
gualism.” The Bangor Centre’s theoretical focus is to understand “the
nature of the relationship between the two languages of bilingual speak-
ers in bilingual communities.”
A useful encapsulation of findings from the past two decades of work is
provided in the Centre for Research on Bilingualism at Bangor’s website
(2014):
New neuroscientific and experimental studies have revealed far more
processing interaction between the languages of a bilingual than was
previously suspected, even when a speaker is only using one language at
a time. Recent developmental studies have emphasised the positive cog-
nitive effects of knowing more than one language. Work by linguists
shows that the use of two languages in the same conversation does not
happen at random, but is constrained in ways that we are just beginning to
understand. Finally, observational and ethnographic research is now
beginning to provide a holistic perspective on the use of two languages
and literacies in interaction at home and school.
In light of the increasing visibility of psychological research on bilin-

gualism, it is instructive to ask whether the perspectives first advanced by
Cook and Grosjean have made a difference to how bilingualism is being
studied in cognitive psychology and/or psycholinguistics. As psychologists
whose own research trajectory in bilingualism began in the late 1980s, and
who have contributed in various ways to the field, we believe it is instruc-
tive at this juncture to reflect on bilingualism research as a whole with
particular attention to the assumptions that pervade this field. We there-
fore sought to examine whether the notion of multi-competence has had a
discernible impact on how bilingualism has been conceptualized and on
how research questions or rationales for research on bilingualism are
framed.
As articulated by Cook (Chapter 1), the multi-competence view has three
central premises: (1) multi-competence involves the total language system
in a single mind (however many languages there may be), (2) multi-com-
petence is important to study in its own right rather than for its bearing on
the monolingual native speaker, and (3) multi-competence affects the
whole mind rather than language alone. If these premises underlying
multi-competence have influenced researchers in the cognitive or psycho-
linguistic study of bilingualism, one would expect that the study of bilin-
gualism would be increasingly framed as having theoretical importance in
its own right. One would also expect a shift away from comparisons
between bilinguals and monolinguals in which monolingual performance
is held to be the normative standard. Finally, one would expect a broad-
ening of the scope of research to include non-linguistic (not just linguistic)
aspects of mental functioning among bilinguals and multilinguals.

Our focus was on the first two issues. In what follows we first provide
evidence bearing on the question of how bilingualism has been influenced
by the multi-competence framework advocated by Cook (1991, 1992). We
conclude that – at least in terms of the criteria we considered – the
influence does not seem to be substantive. We then close by exploring
what a psycholinguistic approach to bilingualism would look like if it were
to take seriously the critiques articulated by Cook and Grosjean over
twenty-five years ago.
4.2 Method
To address the question of the impact of the multi-competence view, we

relied on a combination of quantitative and qualitative observations. With
respect to the former, we undertook a citation analysis of Cook (1991) and
Cook (1992) to see how many of the citations of these works appeared in
empirical psychology journal outlets. We also did a systematic comparison
of empirical studies on bilingualism that appeared in the five-year period
of 1998–2002 and in the most recent five years (2009–2013) to see
(1) whether there has been a change in how studies are being designed
and (2) whether the change is in the direction one would expect if the
multi-competence view has made a difference. In terms of qualitative obser-
vations, we examined how the study of bilingualism is framed and charac-
terized in such venues as conference websites and journal article titles.
4.3 Citation impact analysis
Using Google Scholar we first examined all of the listed citations for Cook
(1991) and Cook (1992). The former reference is where Cook first brings up
the topic of multi-competence and the 1992 reference cites evidence in
support of the notion. For each reference we determined the overall
number of citations of the articles and first classified them by outlet type
(i.e. book or journal). We then classified the journal citations by subject
matter: psychology of language, applied linguistics, education, other.
We conducted our search on July 29, 2014 for Cook (1991) and on
August 1, 2014 for Cook (1992). From the total set of citations listed we
excluded unpublished documents (e.g. drafts of talks posted online) and
self-citations by Cook. This conservative approach resulted in a corpus of
162 citations of Cook (1991) and 344 citations of Cook (1992). From this
corpus, we determined that 32% of the Cook (1991) citations and 35.5% of
Cook (1992) appeared in books, and the remainder appeared in journal
outlets (n = 110 [68%] and n = 222 [65.5%], respectively, for 1991 and 1992).
Given the comprehensive and often interdisciplinary scope of books, we
did not further classify the book outlets by subject matter.

Table 4.1 Summary of journal outlets in psychology of language citing Cook

(1991) and Cook (1992)
Outlet Citations
Cook (1991) Cook (1992)
Applied Psycholinguistics 1 −
Bilingualism: Language and Cognition 1 4
International Journal of Bilingualism 2 8
International Journal of Multilingualism 2 7
Journal of Memory and Language − 1
Journal of Neurolinguistics − 1
Journal of Phonetics − 1
Language Awareness 1 3
Language and Cognitive Processes − 1
Language Learning 1 2
PLOS One − 1
Second Language Research 3 2
Studies in Second Language Acquisition − 1
Of the journal articles that cited Cook (1991) and Cook (1992), we found
that the vast majority were in applied linguistics, sociolinguistics or educa-
tion outlets. For Cook (1991), only 10% of journal citations were in outlets
that cover empirical research in the psychology of language, broadly defined;
for Cook (1992), 14.4% were in such outlets (see Table 4.1 for a summary).
Although it could be argued that it is difficult to interpret these figures
in the absence of baseline figures for cross-discipline citations (e.g. the
extent to which articles appearing in applied linguistics outlets tend to be
cited in psychologically oriented bilingualism outlets and vice versa), the
fairly modest number of citations in psychology outlets – taken as a crude
proxy measure of relative impact – suggests that the seminal papers by
Cook on the notion of multi-competence as an alternative framework for
conceptualizing bilingualism have not had a direct impact on psychologi-
cal research in bilingualism, at least as measured by actual citations of
Cook’s work. Nevertheless, it is possible that there may be an indirect
influence that can be assessed by considering how the topic has been
framed and how studies have been designed. We turn to these issues next.
4.4 Corpus analysis of empirical studies of bilingualism
We identified the total number of empirical studies on bilinguals that

appeared in peer-reviewed journals during two five-year time periods:
1998–2002 and 2009–2013. The first time period was chosen to represent
a five-year period within which any impact of the Cook (1991, 1992) works
or the Grosjean (1985, 1989) articles would have had sufficient time to be
discernible. The later time period was chosen to represent the most recent

Table 4.2 Corpus analysis for the impact of Cook (1991, 1992) and Grosjean
(1985, 1989) using ProQuest. The table summarizes the relative change in
focus on and characterization of bilingualism. N indicates the number of
studies identified in a given time period
Time period
1983–1997 1998–2002 2009–2013
“Bilingual” in title + “monolingual” N = 26 N = 88

in abstract
Biling. versus monoling. com- 50.0% 75.5%
parison groups
Only bilinguals studied N = 15 N = 23
Biling. versus biling. comparison 06.7% 39.1%
groups
The typical bilingual
Equals the early bilingual 61.7% 14.0%
Equals the second language 93.3% 86.0%
user
Linguistic diversity
English dominant for bilingual 69.2% 95.5%
English dominant for 66.6% 74.7%
monolingual
1998–2002 and 2009–2013

1983–1997 combined
“Bilingual” in title N = 66 N = 115

Plus code-switching 10.6% >20%
Code-switching per se N = 22 N = 48
Psycholinguistic perspective 18.2% 28.5%
five-year period. Criteria for inclusion and exclusion of studies were as

follows. We undertook a search of all journals listed in the ProQuest
database that had “bilingual” in the article title and “monolingual” in
the abstract, and we excluded studies of children or of brain-impaired
adults. Based on these criteria we identified a total of 26 studies for the
1998–2002 time period and 88 studies in the 2009–2013 time period. See
Table 4.2 for a summary of our findings.
Bilingual versus monolingual comparison groups. If the multi-
competence view made a difference in the framing of research questions,
one might expect fewer studies in the more recent period adopting a
monolingual group as the standard for comparison. However, a counter-
vailing tendency may also be in effect given that an increased interest in
studying cognitive consequences of bilinguality would lead one to expect
an increasing number of bilingual/monolingual comparisons. With this
caveat in mind, we note that, whereas in the early time period (1998–
2002) 50% of the studies included both bilingual and monolingual
groups, in the recent time period (2009–2013), this percentage has
increased to 75.5% of the studies.

More detailed analysis than was possible here would be necessary to

establish if studies that compared bilinguals with monolinguals were
interpreting bilinguals’ performance in terms of a monolingual standard.
We note parenthetically that in at least two cases – both in the recent time
period – the phrasing of the results in the study abstract clearly indicated
the use of a monolingual standard. For example, in one case bilinguals
were described as producing “less typical responses” on a word association
task relative to monolinguals’ responses; in the other study bilinguals’
performance was described as “deviations from native performance.”
While this suggests that even in the present period the use of the mono-
lingual as the standard for comparison is uncritically held at least by
some researchers, the relative increase in the frequency of studies that
compare bilinguals with monolinguals in itself does not counter the multi-
competence perspective, given that this perspective in fact argues that
bilinguals (and multilinguals) are cognitively different from monolinguals.
Furthermore, as Cook (personal communication, 2014) points out, it is
legitimate – and sometimes illuminating – to compare anything with any-
thing, but it is not legitimate to use one part of the comparison as the norm
and to treat the other as a deviation.
Bilingual versus bilingual comparison groups. If the multi-
competence and holistic views of bilingualism have made an impact
on the psychological study of bilingualism, one would expect to find
more studies over time that systematically compare bilingual sub-
groups varying in some dimension. That is, if monolinguals are not
to be considered the normative standard against which to compare
bilinguals, one would expect that studies of bilingualism inspired by
a holistic or multi-competence perspective would seek to compare the
performance of certain subgroups of bilinguals with other subgroups
of bilinguals. Further, one would expect an interest in examining the
dynamic inter-relationship of the two languages of the bilinguals,
e.g., with attention to bidirectional influences (rather than just the
influence of a first language on a later learned language), and in the
processing of language presented under bilingual presentation or
response conditions, rather than monolingual conditions.
However, when we coded for studies that compared bilinguals with other
bilinguals, we found that these constituted only a fraction of the overall
sample. Specifically, in the 1998–2002 period only one study out of a total of
15 studies examining bilinguals only (or 6.7%) was of this type; in the 2009–
2013 period, 9 out of 23 studies of bilinguals only (39.1%) compared bilingual
subgroups with other bilingual subgroups. Thus it would appear that, for the
majority of cases when bilinguals only are studied, the comparison made is a
within-subjects one (L1 versus L2) rather than a between-subjects
one (bilingual X versus bilingual Y on either or both languages).
Furthermore, in only a few cases (less than 5) did we find comparisons
between monolingual versus bilingual stimulus presentation conditions in
studies of bilinguals.

The preceding analysis suggests that there is clearly room for studies
that compare bilingual subgroups with each other, varying in dimensions
other than proficiency. Possible dimensions could include immigration
history, extent of prior informal translation experience, literacy or
typological distance between different language pairs known/spoken.
The typical bilingual. To determine if studies of bilinguals have con-
sidered the full spectrum of individuals with multiple language experience
or have tended to construct the category “bilingual” more narrowly, we
coded for modifiers such as “early” or “sequential” in describing the
bilingual samples tested. Our findings indicate that the predominant
bilingual type in empirical studies in psychology is the person who
acquired the two languages in sequence, rather than simultaneously.
When considering studies that compared bilinguals with monolinguals,
the bilinguals were classified as early bilinguals in only 1 out of 15 studies
(6.7%) in the 1998–2002 time period and 10 out of 71 studies (14%) in the
2009–2013 time period. Thus, although there is a slight shift towards
greater inclusion of early bilinguals, the majority of studies of bilinguals
continue to construct the typical bilingual as being a late (second lan-
guage) user and, as such, misrepresent what is likely to be the more
common profile globally, at least in multilingual regions such as the
Indian or African continent, where not just bilingualism but early bi/multi-
lingualism is likely the norm.
We also examined the range of languages studied in the two time
periods to get an idea of the linguistic diversity of the populations. We
found that English was by far the most frequently used language of
monolinguals and one of the languages of bilinguals across both time
periods. Specifically, English was the language of monolinguals in 66.6%
of the bilingualism studies in which monolinguals were included in the
first time period (n = 21 studies) and in 74.7% of the studies in the second
time period (n = 79 studies). English was one of the languages of bilin-
guals in 69.2% of the studies in the first time period (n = 26 studies) and in
95.5% of the studies in the second time period (n = 88 studies). Where
monolingual speakers of other languages were studied they included the
following languages: Spanish, French, Swedish, Afrikaans, Catalan,
French, and Quechua. The other languages of the bilinguals studied
included all of the above languages, as well as (typically one or two
studies) involving Italian, Korean, Hebrew, Welsh, Dutch, Tagalog,
Finnish, Chinese, Japanese, and Punjabi. Interestingly, the vast majority
of the studies were conducted in regions where monolingual speakers
are in the majority.
Taken together, we find that there is surprisingly little representation of
languages and language acquisition histories that are likely to characterize
actual bilingual/multilingual users. The canonical profile of the bilingual
in bilingualism studies in psychology is that of the late bilingual for whom
English is typically one of the languages.

Focus on code-switching. If the multi-competence view and the

associated holistic perspective of bilingualism have made a difference
to psychological research on bilingualism, one would expect – over
time – greater research attention to the phenomenon of code-switching,
particularly from a psycholinguistic perspective (see Sridhar and Sridhar
1980 for an early articulation of the need for a psycholinguistic
approach to code-switching and code mixing). To examine this issue,
we compiled a corpus of studies in which “code-switching” was men-
tioned in the title of the article and divided these into two categories:
those that appeared between 1983 and 1997 (that is, before the Cook and
Grosjean articles would have had time to make an impact) and those
that appeared subsequent to their publication.
There were 66 articles from 1983 to 1997 in which the word “bilin-
gual” appeared in the title. Of these, 10.6% also included the term
“code-switching” in the title. In the more recent time frame (based on
the combined total number of articles in the periods 1998–2002 and
2009–2013), there were 115 articles that featured “bilingual” in their
title; of these over 20% also included the term “code-switching.” This
pattern suggests that – while still constituting a minority of studies –
the relative proportion of studies focusing on code-switching has
doubled since 1997.
Next we compared the relative frequency of mention of “code-
switching” per se in the titles of empirical journal articles (regardless
of whether they also included the term “bilingual”). Again we com-
pared two time periods: articles that appeared between 1983 and
1997, and those that appeared in the two time periods noted between
1998 and 2013. A total of 22 articles mentioned code-switching in the
early time period as compared to 48 articles in the later period. This
pattern again suggests an increased research attention to the phenom-
enon of code-switching.
We further classified these latter articles into three categories:
when code-switching was considered primarily (1) in a linguistic/socio-
linguistic and/or teaching/learning context, (2) in a clinical context (as
in switching languages in counseling) or (3) in a psycholinguistic
(language processing) context. Across both time periods examined,
the vast majority of articles featuring code-switching in the title
examined it from a linguistic, sociolinguistic, and/or educational
perspective. In the period from 1983 to 1997, only 18.2% of the articles
on code-switching focused on it from a psycholinguistic perspective
(e.g. looking at the perception or processing of code-switched utter-
ances), whereas in the two time periods from 1998 to 2013, 28.5% of
studies of code-switching examined it from a psycholinguistic
perspective. These observations suggest that psycholinguistic
approaches to code-switching are on the rise: however, they still
manifestly do not constitute the predominant approach.

4.5 Discursive treatment of bilingualism in contemporary

psychological research
Finally, we looked at the discursive framing of bilingualism as revealed in

the language of website descriptions of conferences on bilingualism and in
descriptions of authored or edited volumes on bilingualism. Because we do
not wish to single out particular researchers’ use of these tropes, and
because we do not believe or wish to suggest that they are held by certain
researchers more than others, we will give illustrative examples only
without making reference to specific instances. The tropes that recur in
the language used to frame bilingualism research in the venues we con-
sulted include the notion that bilingualism is a growing phenomenon, that
bilingual language processing is a mystery, and that the brain must do
additional work to “handle” two languages, the assumption here being
that the default case is that of one brain and one language.
These tropes reinforce the notion that bilingualism is a new phenom-
enon (and one on the rise and as such cannot be ignored) and that, from a
cognitive perspective, the interesting questions involve how the two lan-
guages are “managed” or “negotiated.” Use of such language implies
that bilingualism is unusual, intriguing, but also cognitively deviant and
effortful or taxing. Indeed, the notion of “effort” underlies many contem-
porary framings of what bilinguals do: bilinguals are variously described
as “coping” with two languages, or as mental “jugglers” who have to
“handle” two languages. Furthermore, evaluative comments like “remark-
able” or “unusual” or “a feat” in describing what bilinguals do also serves
to emphasize that what they do is something out of the ordinary. In part
the use of such terms may serve to redress previous – typically negative –
depictions of bilingualism (see Hakuta 1986 for such examples from the
early history of bilingualism research in the US).
Similar notions are evident in reviews of the cognitive consequences of
bilingualism. The trope of “competition” is typically invoked – particularly
in its current manifestation of cognitive control – as a need to “inhibit” one
of the languages in production. Instead of such a metaphor, which rein-
forces the idea of a struggle, an alternative conceptualization of the nature
of the “bilingual cognitive advantage” could be in terms of a heightened
vigilance. This latter conceptualization does not invoke a trope of two
language systems in competition with one another.
In summary, the underlying thrust of contemporary ways of framing
bilingualism is to cast the ability to use two or more languages as some-
thing that is remarkable (which implies that it is out of the ordinary).
At the same time this extraordinary ability is also seen as something that
the brain/mind is hard-wired to take on the challenge for. The juxtaposi-
tion of these two views may reflect a gradually emerging realization,
perhaps, that bilinguality or polyglotism is the default.

4.6 Rationales provided for why the study of bilinguals

is important
Based on the multi-competence perspective, we would expect to find a

rationale that states that it is important to study bilingualism because,
despite bilinguals being the global norm, we currently do not know much
about them and it is therefore important to understand bilingualism for its
intrinsic interest. By contrast, a monolingual-focused perspective would
lead one to expect a rationale that positions bilinguals as interesting to
study for what they can tell us about basic cognitive or linguistic
phenomena.
We present the text in full of the description of a symposium on bilin-
gualism held at the 2011 meeting of the American Association for the
Advancement of Science. The author of this text is not stated and our
intent in using the text is primarily by way of illustration – we do not
mean to single it out but would like to suggest that the framing here,
beginning with its title, is representative of how bilingualism is framed in
contemporary (neuro)cognitive research. We italicize words or phrases
that we find relevant in hinting at an underlying perspective.
Crossing borders in language science
What bilinguals tell us about mind and brain

More people in the world are bilingual than monolingual. Historically, the
component disciplines that comprise the language sciences have focused
almost exclusively on monolingual speakers of a single language and largely on
English as the universal language. In the past decade, there has been a
shift in these disciplines to acknowledge the consequences of bilingualism
for characterizing language, understanding the way languages are learned
and used, and identifying the consequences of negotiating life in two lan-
guages for cognitive and brain processes. Recent studies show that bilin-
gualism confers advantages to cognitive control at all stages of life, from
infancy to old age; that contrary to popular belief, being exposed to two
languages from early childhood does not create confusion but instead
modulates the trajectory of language development; that signed and
spoken languages produce a form of bilingualism that is similar to
bilingualism in two spoken languages; and that the continual activity of
both languages affects brain function and structure. Despite the
excitement surrounding these discoveries, we do not understand how
exposure to and use of two languages creates the observed consequences for bilingual
minds and brains. Addressing these questions requires a language science
that is both cross-disciplinary and international. The aim of this sympo-
sium is to illustrate the most exciting of these new discoveries and to begin
to consider their causal basis.
The title and some of the content of the passage above underscore the
second of the two possible rationales – the lessons for monolingualism –
not the one that one would have expected to find if the multi-competence

view were in mind. Similar uses of this frame of “what the study of
bilinguals can tell us about X,” where X refers to some fundamental
cognitive process, are evident in titles of publications by contemporary
researchers in bilingualism. While acknowledging that most people are
bilingual and that knowing two languages does not “create confusion” as
previously thought, the passage nonetheless retains an unquestioned
acceptance of monolingualism as the implicit norm against which bilin-
gualism is to be understood. If bilinguals were truly accepted as the norm
for theorizing language use, it would seem equally important to ask what
happens cognitively when a person fails to become bilingual, or what are
the neurocognitive consequences of not acquiring two languages (see
Cruz-Ferreira 2011 for a similar argument, and also Cook 2009).
Although it is increasingly acknowledged that bilinguals constitute at
least half of the world’s speakers, the main rationale that contemporary
researchers tend to offer for why the study of bilingualism is of theoretical
interest still seems to be what the study of bilingualism can tell us about
basic cognitive processes, rather than what we can understand about
bilingualism itself. The reasoning here seems to be that it is important to
study bilingualism because it provides a unique window into some basic
cognitive process that one could not get by studying monolinguals. While
this framing clearly is designed to make a compelling case for the potential
that the study of bilingualism offers for gaining deeper insight into funda-
mental issues of mental or neurocognitive functioning, it also directs
attention away from bilingualism as a phenomenon worthy of study in
its own right (a central premise of the multi-competence view), which is
what a reverse framing – e.g. “what a focus on attention can tell us about
bilingualism” would have accomplished. Indeed, given that bilinguality is
recognized as more pervasive than monolingualism, the bilingual
language user is, arguably, the norm.
4.7 Conclusion
In this chapter we developed different ways of answering the question of

whether the multi-competence metaphor and/or framework has made a
difference – conceptually and/or methodologically – in how psycholinguis-
tic research on bilingualism has been framed and conducted. We found
that the multi-competence perspective as articulated by Cook (1991, 1992)
has generally been cited infrequently by researchers in psychology.
This “obvious” indicator of impact therefore leads us to the conclusion
that – at least insofar as direct citation is an index of impact – the notion of
multi-competence has not been overtly embraced by researchers of bilin-
gualism in psychology. Of course it could be argued that the particular
outlets in which the two publications by Cook appeared are not ones that
psychologists would normally be in the habit of consulting. Even so, one

could imagine a trickle-down effect – where those who do cite that article
then publish in other outlets that eventually make their way to the outlets
that are more likely to be read by psychologists, so that the ideas expressed
in the original source will eventually make their way. However, for that to
occur one would still expect some kind of crediting or acknowledgment of
the original source, even if it was arrived at through back channels, as it
were. Alternatively, one expects some other kind of influence of the ideas
(not captured by citation frequency) on the ways in which the study of
bilingualism has subsequently been conceptualized and approached.
To get at this issue, we examined the ways in which the study of
bilingualism is framed in contemporary research in psychology, consider-
ing as sources of evidence the language used to characterize it in confer-
ence titles or in abstracts by researchers in the field. One finds four kinds
of framing: bilingualism as something that (1) is cognitively taxing, (2) is a
remarkable feat, (3) has clear cognitive consequences due to the increased
cognitive control requirement, and (4) provides a unique window into
fundamental issues about language and cognition. All these ways of fram-
ing serve to reinforce the notion that single language use is the implicit
norm and that bilingual language use is something unusual, extra, or
special.
These ways of framing the field are echoed in the choice of designs used
by researchers to study the effects of multiple language experience. As
already noted, the predominant comparisons have been between bilin-
guals (or multilinguals) and monolinguals. Bilingual-specific phenomena
(e.g., code-switching or bidirectional language influences) still comprise a
small subset of studies of bilingualism.
Our observations thus lead us to conclude that, although bilingualism as a
topic of inquiry has clearly entered mainstream research in psychology, the
ways in which it is studied and conceptualized appear to perpetuate what
Grosjean termed a “monolingual” or “fractional” perspective rather than a
holistic one. Moreover, despite nearly three decades of theory and research,
the arguments made by Grosjean (1985) and Cook (1991) do not appear to
have substantively affected contemporary research in psychology, except in
certain circumscribed domains of research, such as research on bidirec-
tional influence (e.g. Dewaele and Pavlenko 2003) or on the relation between
language and thought (Cook and Bassetti 2011; Pavlenko 2014). For the
majority of studies in the field, the choice of framing of research questions,
as well as the designs and analyses, do not – either explicitly or implicitly –
adopt a multi-competence framework.
To some extent this relative imperviousness to the holistic/multi-com-
petence perspective is not unexpected given the monolingual/native
speaker foundations on which psycholinguistics as a discipline emerged,
and given the resistance of mainstream cognitive psychology to the study
of individual differences (with the exception of individual differences in
expertise or in age). Nevertheless, we believe it is important to imagine

what a psycholinguistic and/or neurocognitive approach to bilingualism

would look like if multi-competence were to be taken as the starting
premise and monolinguals were constructed as the deviant group
(Cook 1993; Ellis 2006). In the remaining section we engage in such an
exercise.
We would expect that all that we currently take to be a “given” in
psycholinguistic research would have to be recast as applicable only to
“monolingual” language processing. That is, there would have to be a
shift in what is currently the unmarked state of knowledge (Ellis 2006;
see Cook 2009 for an example of what such a shift would look like with
reference to the notion of a multilingual Universal Grammar as the
norm). A consequence of such a shift would be that research on bilinguals
would now become the unmarked form. Monolinguals would be proble-
matized as “not yet bilingual” and a host of new questions could be raised
as to how studying monolinguals could shed light on how language
operates under conditions of deprivation. With respect to studying bilin-
gual populations, we would expect much greater attention to psycholin-
guistic processes associated with (controlled or automatic) moving
between the various languages in the speakers’ repertoire and less atten-
tion to processes involved in single language use, or cognitive control
issues. Furthermore, we would expect greater interest in uncovering and
systematically studying the impact of different types of bilinguals, differ-
ing not only in particular language combinations (e.g. linguistic distance
variable) but also in particular language acquisition histories, i.e. the
varied circumstances (in age and context and motivation and manner
of language acquisition), and much less emphasis on assessment of
bilinguals’ language skills relative to a monolingual or “native” standard.
Thus, we would expect an increase in studies that compare specific
subgroups of bilinguals with each other, rather than studies that com-
pare bilinguals with monolinguals. As part of the emphasis on studying
the impact of heterogeneity, we would expect greater attention to how
language processing changes over time in bilinguals as their linguistic
and extralinguistic environments change. Thus, a dynamic systems
approach is likely to assume even greater prominence (e.g. de Bot and
Makoni 2005; Grosjean and Li 2013; Hernandez 2013).
Further, we would expect bilingual-specific phenomena to become the
object of study to a greater extent than they are currently. Thus, more
studies would be expected that examine the processing and/or retention of
code-switched utterances, as well as studies that examine factors affecting
the production of code-switched utterances in natural settings. More gen-
erally, a greater emphasis on actual patterns of use of their languages in
real-world settings would lead to more interest in the study of the perfor-
mative and creative aspects of language use in bilinguals. As such, we
would expect greater interest in such issues as language play, particularly
playful use involving the juxtaposition of the different languages.

In terms of research designs, we would expect designs that do not

reinforce the notion of translation equivalents by removing words that
are not strictly comparable in meaning across languages, because it would
be recognized that even concrete nouns often do not have exact equiva-
lents (de Groot 2011; Pavlenko 2014). Treating language as an independent
variable serves to reinforce the notion that languages are bounded entities
with clear boundaries, a view that might not fit with how bilinguals
actually perceive, use, and represent words in their languages. Thus,
instead of seeking to compare performance in one language condition
versus another, newer ways of conceptualizing experiments in bilingual
mental processing may explore conditions that are language-open or even
language-optional in their response modes, and may choose to examine
other modes of expression (such as gesture, texting or drawing) concur-
rently with the spoken or written modality.
Studies may turn to exploring language as a dependent variable, to
examine the conditions under which bilinguals may opt to use a parti-
cular language, as opposed to switching between languages or using
their languages interchangeably. Exploring experimental designs in
which participants get to decide on the language(s) of presentation and/
or response rather than letting it be pre-determined by the experimenter
may open up new phenomena that could lead to entirely novel lines of
inquiry on how words/meanings from one language combine and/or
interact with those in another language in ways beyond assumed
“equivalence.” Allusive, aesthetic, cultural or psychosocial links
between the languages may come to be seen as plausible dimensions
along which the mental architecture of the bilingual lexicon might be
constructed (see Vaid and Menter, in press).
Finally, we may expect greater sensitivity to the potential role of extra-
individual influences on language processing (i.e. the influence of cultural
or societal variables that may affect the individual’s feeling of belonging)
and, in turn, how bilinguals perceive and respond to phrases in a particu-
lar language. That is, we may expect more studies that probe the interplay
of affective cultural and sociopragmatic factors on language processing
(see Walters 2004). Studies currently undertaken from a cultural or social
psychology perspective on bicultural identity frame-switching might be
fruitfully integrated with psycholinguistic designs to examine the inter-
play between cultural self-identification and language on perception, rea-
soning, decisionmaking, and personality.
To conclude, we believe that there are promising avenues suggested by
the multi-competence framework that have yet to be seriously explored in
cognitive or psycholinguistic research. There are promising signs in the
work of certain researchers in psycholinguistics who have begun to take
on some of these lines of inquiry that we outline above. As such, we fully
expect to see more evidence of the impact within psychology of the multi-
competence framework over the next 25 years.

References
Albert, M. and Obler, L. K. 1978. The Bilingual Brain: Neuropsychological and

Neurolinguistic Aspects of Bilingualism. New York: Academic Press.
Altarriba, J. and Heredia, R. (eds) 2008. An Introduction to Bilingualism:
Principles and Practice. Erlbaum.
American Association for the Advancement of Science Invited Symposium
2011. What Bilinguals Tell us about the Mind and Brain. http://aaas
.confex.com/aaas/2011/webprogram/Session2808.html
Basque Center on Cognition, Brain and Language. www.bcbl.eu/the-center/
mission/ May 21, 2014.
Berko Gleason, J. and Bernstein Ratner, N. 2012. The Development of
Language, eighth edition. Allyn and Bacon Communication Sciences
and Disorders.
Carroll, D. 2008. Psychology of Language, fifth edition. Stamford, CT: Cengage
Learning.
Centre for Research on Bilingualism, Bangor, Wales. Website. May 21,
2014. www.bangor.ac.uk/bilingualism/centre/index.php.en?subid=0
Chomsky, N. 1959. ‘A review of B. F. Skinner’s Verbal Behavior’, Language 35
(1), 26–59.
Cook, V. J. 1991. ‘The poverty of the stimulus argument and multi-compe-
tence’, Second Language Research 7 (2), 103–117.
Cook, V. J. 1992. ‘Evidence for multi-competence’, Language Learning 42 (4),
557–591.
Cook, V. J. 1993. ‘Wholistic multi-competence: jeu d’esprit or paradigm
shift?’, in B. Kettemann and W. Wieden (eds), Proceedings of the First
EUROSLA Annual Conference, September 1991, Salzburg, pp. 2–8. Tübingen:
Narr.
Cook, V. J. 2009. ‘Multilingual Universal Grammar as the norm’, in I. Leung
(ed.), Third Language Acquisition and Universal Grammar, pp. 55–70. Bristol:
Cruz-Ferreira, M. 2011. ‘The effects of monolingualism’, Being
Multilingual blog posting. http://beingmultilingual.blogspot.com/
2011/07/
De Bot, K. and Makoni, S. 2005. Language and Aging in Multilingual Contexts.
An Introduction. New York: Psychology Press.
De Groot, A. M. B. and Kroll, J. F. 1997. Tutorials in Bilingualism: Psycholinguistic
Perspectives. New York: Psychology Press.
Dewaele, J.-M. and Pavlenko, A. 2003. ‘Productivity and lexical diversity
in native and non-native speech: a study of cross-cultural effects’, in

V. J. Cook (ed.), Effects of the Second Language on the First, pp. 120–141.
Dil, A. (ed.) 1972. Language, Psychology and Culture: Essays by Wallace E. Lambert.
Stanford, CA: Stanford University Press.
Ellis, E. 2006. ‘Monolingualism: the unmarked case’, Estudios de
Sociolingüı́stica 7 (2), 173–196.
Ervin, S. and Osgood, C. 1954. ‘Second language learning and bilingual-
ism’, Journal of Abnormal and Social Psychology 49, 139–146.
Evans, N. 2013. Multilingualism as the primal human condition: What we
have to learn from small-scale speech communities. Keynote Address,
International Symposium on Bilingualism, Singapore, June 10–13.
Fernandez, E. M. and Cairns, H. S. 2010. Fundamentals of Psycholinguistics.
New York: Wiley-Blackwell.
tual view’, Modern Language Journal 95, 344–355.
Gardner, H. 1985. The Mind’s New Science: A History of the Cognitive Revolution.
Boston: Harvard University Press.
Grosjean, F. 1982. Life with Two Languages. Cambridge, MA: Harvard
University Press.
Grosjean, F. 1985. ‘The bilingual as a competent but specific speaker–
hearer’, Journal of Multilingual and Multicultural Development 6 (6),
467–477.
Grosjean, F. 1989. ‘Neurolinguists beware! The bilingual is not two mono-
linguals in one person’, Brain and Language 36 (1), 3–15.
Grosjean, F. and Li, P. 2013. The Psycholinguistics of Bilingualism. Wiley-
Blackwell.
Hakuta, K. 1986. Mirror of Language: The Debate on Bilingualism. New York:
Basic Books.
Harris, R. (ed.) 1992. Cognitive Processing in Bilinguals. Amsterdam: North
Holland.
Hernandez, A. 2013. The Bilingual Brain. Oxford University Press.
International Conference on Multilingualism: Linguistic Challenges and
Neurocognitive Mechanisms. http://multilingualism.conference.mcgill
.ca/invited-speakers.php
Keatley, C. W. 1992. ‘History of bilingualism research in cognitive psychol-
ogy’, in R. Harris (ed.), pp. 15–50.
Kecskes, I. 2010. ‘Dual and multilanguage systems’, International Journal of
Multilingualism 7 (2), 91–109.
Kroll, J. F. and de Groot, A. M. B. (eds) 2009. Handbook of Bilingualism:
Psycholinguistic Approaches. Oxford University Press.
Lambert, W. E. 1977. ‘The effects of bilingualism on the individual: cogni-
tive and sociocultural consequences’, in P. Hornby (ed.), Bilingualism:
Psychological, Social, and Educational Implications, pp. 15–27. New York:
Academic Press.
Li Wei (ed.) 2000. The Bilingualism Reader. New York: Routledge.

Osgood, C. and Sebeok, T. 1954. Psycholinguistics: A Survey of Theory and

Research Problems. Baltimore: Waverly Press.
Paivio, A. and Begg, I. 1981. Psychology of Language. Englewood Cliffs, NJ:
Prentice-Hall.
Paradis, M. 1980. Readings on Aphasia in Bilinguals and Polyglots. Montreal:
Didier.
Pavlenko, A. 2002. ‘“We have room for but one language here”: language
and national identity in the US at the turn of the 20th century’,
Multilingua 21, 163–196.
Pavlenko, A. 2014. The Bilingual Mind. Cambridge University Press.
Reynolds, A. (ed.) 1991. Bilingualism, Multiculturalism and Second Language
Learning: The McGill Conference in Honour of Wallace E. Lambert. New York:
Psychology Press.
Sridhar, S. N. and Sridhar, K. K. 1980. ‘The syntax and psycholinguistics of
bilingual code mixing’, Canadian Journal of Psychology 34, 407–416.
Taylor, I. and Taylor, M. 1990. Psycholinguistics: Learning and Using Language.
Englewood Cliffs, NJ: Prentice-Hall.
Vaid, J. (ed.) 1986/2014. Language Processing in Bilinguals: Psycholinguistic and
Neuropsychological Perspectives. Hillsdale, NJ: Lawrence Erlbaum. 2014 edi-
tion reprinted by Taylor and Francis.
Vaid, J. 2006. ‘Joking across languages: perspectives on humor, emotion
and bilingualism’, in A. Pavlenko (ed.), Bilingual Minds: Emotional
Experience, Expression, and Representation, pp. 152–182. Clevedon:
Vaid, J. and Menter, R., in press. ‘Languages without borders: reframing the
study of the bilingual mind’, in G. Libben, M. Goral and M. Libben (eds),
Bilingualism: A Framework for Understanding the Mental Lexicon. Amsterdam:
John Benjamins.
Vaid, J., Paivio, A., Gardner, R. C. and Genesee, F. 2010. ‘Wallace E. Lambert
(1922–2009)’, American Psychologist 65 (4), 290–291.
Walters, J. 2004. Bilingualism: The Sociopragmatic–Psycholinguistic Interface.
Warren, P. 2013. Introducing Psycholinguistics. Cambridge University Press.
Weinreich, U. 1953. Languages in Contact: Findings and Problems. New York:
Linguistic Circle of New York.
Yildiz, Y. 2012. Beyond the Mother Tongue: The Post-Monolingual Condition. New
York: Fordham University Press.

5
Multilingualism research
Rita Franceschini
The emergent field of multilingual research has had a much greater impact
on linguistics than many might have thought possible. The research subject
of ‘more than one language in one mind’, for example, represents not
merely an additive function, but rather a separate challenge altogether,
thus breaking the boundaries of traditional methods of description. This
and other findings have shaken some of the very foundations of linguistics:
it affects for example the view of clearly definable languages with their
modules, and ultimately the separability of languages in the mind of an
individual. When the focus is on the multilingual individual, the concept of
separable individual languages can no longer be strictly applied.
Multilingualism research introduced a much-needed element of
renewal in the field of linguistics, which had been, albeit not explicitly,
bound to a romanticized single-language view despite some structural
renewal and expansions throughout the last century: the social turn (with
the dawn of sociolinguistics), the pragmatic turn as well as the spatial turn
were and are important steps that keep reverberating through modern-day
research, in the field of sociolinguistics (which happens to be greatly
influenced by bilingualism and language contact research à la Weinreich
(1953) as well as in pragmatics and language acquisition research, to name
but a few fields that this chapter will address.
Today, a new view is forming – one that no longer exclusively focuses on
the structure, but rather on the underlying processes that result in coagu-
lated patterns of structure and constructions with a strong cultural and
communicative imprint. The view of language use in a specific action
context, which dates back to the beginnings of the twentieth century
(think of Ludwig Wittgenstein’s Philosophical Investigations, the re-discovery
of Vygotsky), keeps spreading. Accordingly, language is seen as embedded
in a context of socially situated action and linguistic knowledge as well as
the underlying cognition and abilities that are closely linked to these social
experiences.

98 R I TA F R A N C E S C H I N I
5.1 The multilingual individual emerges on the horizon . . .
Before Weinreich’s research (1953), linguists’ knowledge of the multilin-

gual individual was fragmentary at best. If at all recognized, such indivi-
duals were labelled as rare and special cases that represented an enigma
against the backdrop of common assumptions in linguistic theories. This
individual did not fit into any one category, and had such inexplicable
habits as switching from one language to another – sometimes even in one
and the same utterance. The changed, increasingly open and interlingual
approach shed light on a wide variety of cross-linguistic patterns all over
the world, and despite the fact that bilingualism and multilingualism had
been practised since time immemorial (for the Roman Empire see Adams
et al. 2002), it was only then that research started to look into such
phenomena: code-switching and code mixing eventually became some of
the most thoroughly analysed phenomena in bilingualism research
(Gardner-Chloros 1991; Myers-Scotton 1993; Auer 1998).
The emergence of a more thorough analysis of people who use more
than one language in everyday life also created a need for new terminology
to accurately describe the new observations. This new terminology cannot
be seen only as a new, short-lived trend: it indicates the need to concep-
tualize new knowledge that was made possible by new approaches and
data collections. New technology such as audio and video recordings and
corpus linguistics introduced over the last decades gave researchers new
tools that allowed them, for example, to do micro-analytic studies in
conversation analysis. The new tools allowed for a level of detail in
research that was simply impossible to achieve before the digital era. As
a result, many corpus linguistics and data-related analyses gained a strong
foothold alongside experimental approaches that were and are mainly
used in psycholinguistics. In addition, and not only since the proclamation
of the ‘decade of the brain’, new psychometrical and neurobiological
research into the ‘ bilingual brain’ and the ‘multilingual brain’ has entered
the equation.
When linguistics started to look more closely into language use and
derived new knowledge from these observations – and especially since
the focus has turned to achievements of people who regularly use more
than just one language – a need for new concepts emerged. Since
Weinreich’s groundbreaking work (Weinreich 1953), many concepts
have been successfully introduced, including interlanguage, interference,
transfer, endolingual and exolingual behaviour, language choice and shift,
repertoire, code-switching, code mixing, fused lects, crossing, bricolage,
translanguaging, superdiversity, linguodiversity, code-meshing, continua
of biliteracy, polylingual languaging . . . 1 It is not always evident whether
each term indicates the product of multilingual speech, the process involved,
or both. At any rate, the new terminology aims at classifying linguistic
cross-over phenomena in a variety of ways and from different perspectives.

Multilingualism research 99
The criticism followed swiftly: some of the terms assume – consciously or

unconsciously – that there is such a phenomenon as ‘clearly separable
languages’; that a speaker can ‘switch’ from one language to the other; or
that expressions can be analysed and subsequently classified as belonging
exclusively to ‘one’ or ‘another’ language. A general experience is that
such a differentiation often fails when bilingual speech is more deeply
analysed (Franceschini 1998). The difficulty lies in the definition of ‘lan-
guage’ (for the ‘six meanings’ applied to language learning see Cook 2007)
or, according to Hall et al., in the notion of ‘the existence of autonomous
systems with identifiable borders’ (Hall et al. 2006, p. 231).
The proliferation of terms raises a warning flag: it either indicates the
existence of multiple interlingual patterns where unity has yet to be
achieved (everyone coins their own term and tries to establish it in
the scientific world), or shows that the phenomena are so diverse that
the situation truly demands such a large number of terms to adequately
give order in an enormous field of different phenomena. In other words, is
the proliferation of terms based on a general heuristic concern or does it
genuinely reflect a high level of variation?
To sum up: there is still an immense description gap between what a
speaker does and knows and the manner in which linguists describe and
classify the speaker’s behaviour. Often, it seems that we oversimplify
matters in our work: we are standing in our own way if we continue to
use the traditional method of description that focuses on individual lan-
guages. The multilingual speaker does not follow us: he or she is way ahead
of us.
5.2 The term multi-competence emerges

on the horizon . . .
The term multi-competence (hereinafter MC) emerged at a time (Cook

1991, p. 112; Cook 1992) when research became increasingly aware of
the ‘explanation gap’ among:
– psycholinguistic approaches
– cognitive perception, and
– usage-based approaches.
The absence of connections among these three approaches became all the
more evident when researchers began to analyse the linguistic knowledge
of individuals who use more than one language (for an overview see
Ortega 2011).
It would go beyond the remit of this chapter to trace the full history
of the term MC (Kecskes 1998; Alptekin 2010; Franceschini 2011a), and I
will limit myself to the following: the generativist paradigm proved to
be insufficient in efforts to model the co-presence of two – or

more – languages and their respective grammar in one speaker. With the
creation of the term and its further development, Vivian Cook overcame
this paradigm by labelling the competence of a bilingual speaker as a
describable and possible competence ‘in its own right’ (Cook 1992), rather
than a deviation from ‘native’ competence. And the competence of bilin-
gual speakers – including its characteristics and principles, which differ
from those of a ‘native speaker’ – requires a description of its own. It
should not be measured against the competence of an ideal, native
speaker/hearer: the learner is not perpetually a defective speaker, but
has their own competence with characteristics that need to be researched.
In retrospect, the term MC did not originally gain its importance simply
by virtue of its emergence. In fact, its original definition was very short:
‘The compound state of a mind with two grammars’ (Cook 1991, p. 112).
First and foremost, the significance of the term’s success and its expansion
lay in the fact that it represented a liberation that attracted linguists of
different provenance: the term allowed for a more holistic approach to
competence that included all influences regardless of their influence and
direction (from L1 to L2, from L2 to L1). The term matched the zeitgeist of a
period that was inspiring for many who were interested in multilingual
phenomena. The brief definition allowed for various interpretations,
which had the positive side-effect of giving rise to an entire movement.
Starting from the above-mentioned brief description of MC, the use
of the term became increasingly compatible with sociolinguistic
environments (Franceschini 2011a), with the definition – also for this
book – today reading as follows: ‘The overall system of a mind or a
community that uses more than one language’ (Cook, Chapter 1).
5.3 Language biographies or the complexity of the

experience with one’s own languages
I first started being interested in using the term MC in the late 1990s in
connection with the analysis of language biographies. In our terms,
language biographies are narrative autobiographies that use the method of
in-depth interviews (see Schütze 1987). The narrative-generating question
was: ‘How did you come to use your different languages?’ In other words,
the interviews focused on the topic of languages: the floor maximally open
for the narrator, with the interviewer assuming the role of an attentive
listener who gives feedback and only asks further questions once the
interviewee seems to have finished. Our recordings were made in a
Central European context (interviews in German, Romance and Slavic
varieties), more specifically in Basel, the overall Alemannic dialectal area
in Switzerland, the Czech–Slovenian and the Northern Italian regions
(mainly Trentino–Alto Adige/Südtirol), with data collected primarily in
the Basel Prague Project (BPP) and Multilingualbrain projects as well as in

other contexts.2 On average, the interviews are one-and-a-half hours long,

with the corpus now consisting of more than a hundred interviews.
It soon became apparent that the ways in which languages are acquired
are much more complex than was generally assumed in second language
acquisition (SLA) research (see also Kramsch 2009). Language acquisition –
both untutored and tutored – turned out to be a very intricate matter,
closely linked to personal experience, the speaker’s attitude toward socie-
ties and the narrow social context. In brief, it is a very individual experi-
ence. Humans are, after all, social beings, and acquire languages in and
through social contexts.3 As a result, the speaker’s utterances reflect, both
formally and in terms of content, traces of everything that has shaped his
or her individually experienced behaviour: the phonology indicates the
speaker’s regional and social roots, while the grammar, syntactic patterns
or use of rhetorical questions can give hints regarding the language acqui-
sition context. The way a story is told mirrors the speaker’s experience of
contact and interaction with other individuals and groups or an entire
society throughout his or her life.
Autobiographical narration – here focusing on language acquisition – is
portrayed by the narrator as a destiny, a life experience and an individual
achievement; and a comparison of many such accounts shows how closely
intertwined these personal narratives are with collective experiences. This
explains why narration patterns and motives are repeated throughout a
region, a community, or a group.
This interconnection is a reflection of the fact that languages and their
acquisition take place in a social environment, that they impact a socio-
historical context and are in turn also affected by this context. Societies
and groups with their individual way of dealing with languages, dialects,
and register – in other words, with language varieties – are reflected in all
their diversity in language biographies: take, for example, societies and
groups that lived mainly with natural, historically rooted language contact
as opposed to societies and groups with a monolingual habitus; commu-
nities that are integrative compared to communities that have segregating
tendencies; think of communities where languages and dialects are con-
nected with values of different social stratification and communities
where dialects are used neutrally without reference to social layers
(such as the above-mentioned Alemannic dialects); societies where multi-
lingualism is an everyday, pervasive phenomenon or societies in which
other languages are only used in formal education through instructed
learning. All these aspects become a topic in the narrator’s reflection and
personal interpretation, with the interviewer also becoming a part of the
narration: aspects such as ‘What does the interviewer already know?’ and
‘What do I need to explain?’ are a constituent characteristic of language
biographies. The interviewees display in front of the interviewer the multi-
faceted composition of their individual repertoires, enriched throughout
their lives through contact and regular exchange with other individuals

who themselves can have completely different repertoires. This contact

affects both sides and results in the repertoires thus acquired – consciously
or unconsciously and implicitly. This is why every individual carries
within him or her specific traces of his or her linguistic experiences.
A recurring characteristic in language biographies centres on languages
acquired at school: in our language biographies collected among multi-
lingual persons, these languages play a less dominant role compared to the
acquisition of languages and varieties in everyday contexts based on a
direct, interactive exchange with other people. The latter appears to be,
in our data, more important in a speaker’s memory and emotional embed-
ding. Language biographies, therefore, are suitable as a tool to compile and
analyse a ‘repository’ of language contact.
It is the social embedding that explains the existence of recurring pat-
terns and themes in all acquisition processes that we were able to view in
various projects throughout the past nearly 25 years. I would like to high-
light one recurring theme that can shed light on a neglected aspect of MC.
It became clear that language acquisition also takes place by the repeated
and continuous exposure to a social environment marked by various
languages: for example, when other languages are spoken in one’s neigh-
bourhood, when an individual hears other children speak another lan-
guage during childhood, or when there are signs and writings in another
language in the individual’s normal environment.4 The languages of a
social environment are not consciously learned (for example at school
where these languages are often not treated as legitimate languages), but
rather subtly absorbed, with the result of a minimal language acquisition
(that can nevertheless be measured: Franceschini 1999, 2003a, 2011b,
2012). Sometimes this kind of acquisition achieves even more than just
minimal competencies. I have labelled this process unfocused language
acquisition (to distinguish it from implicit learning). During in-depth lan-
guage biography interviews, varieties that are acquired incidentally and in
which the narrator can often merely get by are mentioned very often.
According to our corpus, these varieties have emerged almost exclusively
through contact, mainly with immigrants. Among adolescents, such
effects are often accompanied by emphasis and playful creativity
(Rampton 1995; Dirim and Auer 2004), but basically, such crossing effects
also take place in other groups that are in contact with speakers of other
languages. The phenomenon of a flexible and nonchalant handling across
several languages may be more evident among adolescents, but such
language ‘mixture’ generally indicates a way of life in which contact
with other languages is nothing out of the ordinary but rather a habitual
occurrence, and that the ‘borders’ defined in traditional linguistics may
not apply.5 When we asked people with ‘mixed’ language behaviour
whether they are familiar with language X or Y, we were often met with
incomprehension. The recurring answer was: ‘Well, that’s just how we
talk’ (Franceschini 1998, 1999).

Within the vast field of so-called spontaneous language acquisition,

implicit learning via unfocused language acquisition is neglected: usually
(but not invariably), it is shrugged off as merely marginal under the
assumption that it gives the speaker only a minimal knowledge of the
language (Franceschini 2003a, 2011b, 2012). Well, the exposure to a lan-
guage – often dismissed as ‘passive behaviour’ – definitely leaves its marks.
5.4 Exposure matters!
What matters is the exposure to various languages that are an integral part
of the social context in which a speaker grows up and lives and whose
actions he or she is a part of.
The finding that exposure is an important variable is also the result of
neurobiological studies that we have conducted in an urban context
among young adults (average age 28 years) who are fluent in three
languages.6 They solved a language production task (silent free narration
task, in replication of Kim et al. 1997) while being measured by fMRI
(functional magnetic resonance imaging). The individual and contrastive
analysis of the language biographies showed that the 44 subjects differed
with respect to age and contexts of second language acquisition, but also in
the quantity and quality of the exposure to that second language.
Accordingly, they were classified into four distinctive groups (see Bloch
et al. 2009, pp. 626–628):
A Simultaneous bilinguals. These grew up in a bilingual environment; since

birth they had contact with persons in their close environment who
regularly interacted with the child in two languages.
B Covert simultaneous bilinguals. These were born into a monolingual
family whose language differed from the one spoken in the surround-
ing context. While having only little and irregular direct interactive
contact with this second extra-familial language, they were nonetheless
exposed to it since birth, leading to a ‘passive’ competence that was
later on activated by an increase in input and direct interaction.
C Sequential bilinguals (age of L2 acquisition, 1–5 years): these subjects
were born into a monolingual family speaking the language of their
surrounding environment. Because of the emigration of their family to
a country in which a different language was spoken, they acquired their
L2 between the ages of one and five years.
D Late multilinguals. These subjects were born into a monolingual family
speaking the language of their surrounding environment. These sub-
jects learned their first foreign language at school, i.e. at the age of nine
years or older.
During the first few years of their lives, the subjects of Group B grew up
in a monolingual family whose language differed from the one spoken in

the surrounding context (e.g. growing up in Basel in a Kurdish-speaking

family). They were ‘only’ exposed to L2. Group A, in contrast, grew up in a
bilingual environment where both languages were used in the family. One
could argue that Group B had a less explicit exposure and Group A had a
more direct one: for Group B, the other language was the ‘white noise’ of
their social environment, whereas Group A was exposed to both languages
through everyday interaction with their caregivers. It appears safe to
assume that the language environment in which Group A grew up was
more influential, whereas L2 was clearly an ‘environment language’ for
Group B. Still, there were no significant differences: in their behaviour as
adults, Groups A and B did not differ in the performance of their language
production task. At first glance, this is surprising. How can this phenom-
enon be explained? It is generally accepted that the presence of an L2 has
an impact on the neuronal substrates used for language processing; the
influence of the age of L2 exposure, however, is still controversial.7 In our
study (Bloch et al. 2009), subjects with early exposure to L2 showed low
variability in brain activation in all three languages. In contrast, late multi-
linguals exhibited higher variability. This shows that variance in represen-
tation of a multilingual repertoire increases gradually with the age of L2
acquisition: late multilinguals – i.e. Groups C and D – exhibit much higher
individual variance. The data suggests a gradual change from the uniform
representation of several languages in ‘early multilinguals’ to the more
heterogeneous representation in ‘late multilinguals’. Early exposure to an
L2 (as in Group B) results in the same low variance as active bilingual
upbringing (Group A):
early exposure to two languages results in the formation of a language
processing system that is able to handle all languages acquired, including
late learned languages. Late multilinguals, in contrast, show much more
variability; the different languages cause an inhomogeneous activation.
Apparently, monolingual upbringing results in a language network which
is not able to accommodate late learned languages to the same degree as it
is the case in ‘early multilinguals’. (Bloch et al. 2009, p. 631)
Our data show that:

early covert exposure to a L2 results in a similar effect on variance as overt
bilingual upbringing is doing. Covert Simultaneous subjects who were una-
ware of the exposure to a second language show the same results as the
Simultaneous subjects. In fact, linguistic studies have shown that ‘ passive’
exposure to other languages during childhood can lead to an unfocussed
form of language learning and to a form of competence that can be
reactivated at later stages if necessary. (Bloch et al. 2009, p. 631)
These research findings give one more reason not to neglect ‘margin-
alized’ varieties in an individual repertoire. The language production
study showed that a person’s linguistic environment during childhood
and adolescence has an impact on cognition. In consequence, exposure

and the underlying process of unfocused language acquisition must not

be neglected.
5.5 Multi-competence is the ‘third quality’
Throughout the course of their lives, speakers accumulate experiences in

dealings with a variety of other speakers, various sound environments that
are not necessarily fully understood, visual input that captures the speak-
er’s attention in residential environments without the speaker focusing
consciously on such aspects. All these competences and the awareness of
the variety of linguistic action form the speaker’s repertoire.
And yet, unlike grammar books on a bookshelf, these competences and
the awareness thereof (even if it is ‘only’ practical knowledge) are not
stored next to one another in the speaker’s mind. Rather, they are con-
nected and they build on each other; they even establish new cross-refer-
ences. It is still a matter of debate as to how this works, as indicated by
research into transfer phenomena from L1 to L2 and vice versa (see Cook
2003 as a work of reference). The acquisition of each new language
changes the structure, even that of the L1, in terms of the assignment to
categories and judgments regarding grammar. One thing is certain: new
experiences infiltrate existing ones, are integrated more or less strongly
and can be reactivated for future reference. This is the first significance of
the ‘third quality’: a quality that represents more than the sum of its parts.
The term MC has proven particularly useful here: it helps when referring
to a competence that involves the use of several varieties.8 In addition, in
terms of the language biography research described above, MC includes
the social formation of these competences.
Based on what has been stated so far, an individual’s MC could embrace
at least implicit knowledge or ability regarding:
(a) socially constructed and legitimized, prestigious speech (and writing);

(b) speech (and maybe writing) of varieties that are used in groups without
higher-ranking, legitimized standards (from communities of practice
all the way to marginalized groups);
(c) contextual adequacy of the use of all varieties acquired throughout a
lifetime, including the social and cultural adequacy of the application
thereof;
(d) similarities and contrasts between these uses;
(e) flexibility in the experience of varieties and their use; and/or
(f) ability to combine different varieties and to create neologisms.
It goes without saying that all aspects include both active and receptive
competences. Depending on the range of experience and contact exposure
that a person has accumulated throughout life, the level of knowledge and
ability of these aspects can vary greatly.

This description of competences does not use the term ‘languages’, but
rather operates with the term of varieties (see also note 7). This should
serve as a more neutral means to define language uses that groups have
created as perceptible language patterns that establish their identity over
the course of their communication history. Some of the socially produced
varieties may have received a name of their own (a glottonym), especially
after an expansion that allowed these varieties to embrace many different
functions (within the meaning of Ausbausprachen as introduced by Kloss
1967), or they could have a broader reach and prevail over other rival
varieties. Other varieties may be group varieties with a smaller reach of
action and comprehension or may take the form of a temporary phenom-
enon, as is sometimes the case in youth language varieties.
In a broader sense, the concept of MC is also of interest in multilingual-
ism research and describes the ‘third quality’ of multi-competent speak-
ers: it allows the modification of a speaker who moves about freely and
swiftly within his or her communicative repertoires (see (a)–(f) above) in an
effort to secure understanding, using not only the stronger or more fluent
variety, but also the ones that he or she uses less frequently or tends to
marginalize, or even varieties he or she may only have minimal knowledge
of: thus, a flexible speaker (Franceschini 2003b). This is an essential compo-
nent that the above-mentioned ‘third quality’ of MC can describe: the
ability to use the different varieties flexibly and promptly within one and
the same communication. The multi-competent speaker is not only able to
use single varieties to a full Ausbau, but can also be familiar with elements
of the other communication partner’s full language version. And above all,
a multi-competent speaker has the ability to adequately use his or her
competences and to switch or glide between the individual varieties (e.g.
between standard and regional varieties). In particular, the speaker has the
ability to use different varieties in order to bring about transparency in
between his or her varieties. Let’s look at an example: a speaker who is
familiar with Swiss German (an Alemannic variety), but not with English
(because English only constitutes a marginalized language in the speaker’s
repertoire), will not find it very difficult to grasp the concept of the term
slippery when he or she first encounters it: the Swiss German equivalent is
schliferig. The same holds true for to giggle: giggele is the Swiss German
expression with the same meaning. Hearing the expression atenziun uffants
in the Romantsch-speaking part of Switzerland, French will help you to
understand that children (French enfants) are meant; in the same region,
Latin helps to understand the name of the restaurant Crusch alba (‘white
cross’). In Sardinian, the term la nappedda (a spoon made of one piece of
cork used to lift water) is more transparent for multilingual speakers with
French, because it reminds the listener of napper (‘to sprinkle’), and so on.9
For a multi-competent speaker, making such connections throughout
one’s own repertoire is natural, and he or she uses them to solve commu-
nicative tasks, beginning from comprehension.

Multilingual speakers may be more likely to recognize such cognates or

near-cognates, but the connections between the different varieties may be
more varied, the effects of guessing and trying out may be greater, and as a
result understanding as a whole may be intensified. The likelihood of a
multilingual speaker using these effects is high, and even higher if the
varieties are genetically related.
If this re-recognition potential is used – consciously, but more often
unconsciously – I would ascribe it to the above-mentioned ‘third quality’
that goes beyond a merely additive multilingualism.
Therefore, from the point of view of multilingualism, a multi-competent
person is a speaker who makes flexible use of his or her entire language
capabilities by connecting the varieties dynamically and in a manner that
is adjusted to the demands created by the given communicative situation,
both in reception and production.
A multilingual speaker’s repertoire must therefore be seen as a dynami-
cally growing entity that changes according to the experiences that the
speaker makes, with new competences only seemingly representing an
additive element: as soon as new information is absorbed, it is aligned to,
integrated in and connected with existing knowledge. Without this incor-
poration – or docking – in already existing knowledge, no new knowledge
can be absorbed. Incomprehensible elements remain inscrutable, opaque
and beyond the reach of cognition. However, if a connection can be made
to existing knowledge, integration – and ultimately learning – can take
place, both explicitly and implicitly.
5.6 Multi-competence and repertoire
As is highlighted above, from the point of view of multilingualism, the

term MC is closely connected to the term repertoire. The latter was initially
used on a macro-level to describe languages in use in a society; a similar
use on the individual speaker’s level followed swiftly, as it seemed logical
and simple to refer to the entire range of languages that an individual uses
as his or her personal repertoire (Gal 1987). It seems that this was made
under the premise that languages can be counted and clearly separated.
This possibility cannot entirely be refuted; after all, also from a speaker’s
perspective, social groups may also want to sharply distinguish varieties,
for example in areas where language communities have a desire to keep
separate from one another (e.g. in the former Yugoslavia or in geopolitical
hotspots where government officials are campaigning for separation). The
situation is different, however, when languages are analysed and differ-
entiated from an insider’s emic viewpoint (in the sense of K. L. Pike). It is
possible to interpret languages from the point of view of language use, and
observe them as a creation in individual interactions, as briefly outlined
earlier. Taking this viewpoint, multilingual speakers perform a wide array
of actions that cannot be forced into predefined categories.

If we assume that language use is socially contingent, then we also

assume that the actions are integrated in a social context. We would not
primarily base our observations on the notion of single languages
(although on secondary consideration we might, where appropriate).
This usage-based perspective sees the speaker/listener as an individual
who takes linguistic action and uses linguistic means in order to commu-
nicate adequately. He or she can, as part of this process, use socially
legitimized and clearly outlined varieties that enjoy a high reputation, as
we have seen previously. Yet the same speaker may, shortly after, choose
to use less legitimized uses of language by freely and flexibly mixing
and matching expressions or sequences of other varieties. This practice
has been referred to as bricolage – based on Claude Lévi-Strauss’s work
(Lévi-Strauss 1962).
Compared with the level of flexibility that a term should have in order to
cater to such linguistic behaviour, the term repertoire is rather rigid and
committed to an additive view. Relations and connections between lan-
guages are not sufficiently focused on. What we need is a term that covers
the full range of varieties that a person uses, sometimes more, sometimes
less successfully. The more agile term MC more appropriately describes
the above-mentioned ‘third quality’, which is far from being fully explored
as of yet.
5.7 What counts is variability!
Given a centuries-long tradition of written standardization processes for

languages, some (Euro-centric) language observers may find it difficult to
distinguish between a forcefully produced institutionalized language form
and the situational language use in everyday life, with the latter forming
the ‘basic’ way of life. The varieties of everyday, situational action between
communication partners – I would like to refer to them as usage varieties –
are more variable, as they are used in natural contexts and have to adapt to
permanently changing conditions. In this process, some groups may create
standards that become dominant and will eventually become part of the
standardized norm. Once again, a multi-competent person is aware of the
context-sensitive use of different varieties, ranging from institutionally
standardized varieties, as described above, to group norms and usage
varieties. While some have a broader range than others, all of them use
the full range and can cater to a variety of registers.
The concept of MC is all about covering the use of different varieties in
different situations. As such, the concept of MC would be based on that of
language variability. It is therefore assumed that people are able to be
multi-competent because they are able to move freely within the sphere
of their personal communication experience created in the course of their
lives. But this would suggest that in extreme cases, a person can be labelled

as multi-competent if he or she is able to communicate with a certain

variety even if he or she, viewed from the outside, speaks ‘only’ one
language. Following this logic, it’s not monolingualism or multilingualism
that matters, but rather the general ability of variable language use.
Hall et al. (2006) came to a similar conclusion: instead of speaking about
‘multilingual persons’, they speak of ‘communicative experts’, or more
specifically of speakers that are ‘multi-contextual communicative experts’
(Hall et al. 2006, p. 232). An analysis of my own data – above all of language
biography interviews – prompts the following view: it is communicative
expertise that defines a multilingual person. In some – the so-called mono-
linguals – this expertise is marked by a set of varieties that are generally
ascribed to one ‘language’; in others – based on their communicative
experience – more languages/varieties are involved that are, in a social or
scientific discourse, allocated to different glottonyms. Both kinds of speak-
ers can, however, be flexible and competent. As such, a monolingual
speaker could also be seen as multi-competent.
Given these premises, does it still make sense to maintain multilingual-
ism as a term? Comparing the career of the concept of MC with that of
multilingualism, they gradually grew closer: Multilingualism was a crea-
tion derived from a sociolinguistic point of view, and still takes social
factors into close consideration. It has remained a very comprehensive
term. I have previously proposed to use a definition based on four dimen-
sions that need to be differentiated when using multilingualism, depend-
ing on the point of view taken.
A definition of multilingualism
The term/concept of multilingualism is to be understood as the capacity of
societies, institutions, groups and individuals to engage on a regular basis
in space and time with more than one language in everyday life.
Multilingualism is a product of the fundamental human ability to commu-
nicate in a number of languages. Operational distinctions may then be drawn
between social, institutional, discursive and individual multilingualism.
The term multilingualism is used to designate a phenomenon
embedded in the cultural habits of a specific group, which are charac-
terised by significant inter- and intra-cultural sensitivity.
(Franceschini 2009, pp. 33–34)
The last two dimensions refer to the level of group interaction, in other
words to situated interaction that has gained considerable attention over
the last decades (see the many studies that primarily make use of micro-
analytical methods); the very last dimension focuses on the individual
level, in other words on the contact of languages within an individual
and on acquisition processes. It is here that the concept of MC fits in –
and at the same time, we see that the term MC today also involves social
embedding. The two terms approach one another (see Figure 5.1): MC
transcends – from left to right – the individual view with the definition

Multi-competence Multilingualism
Arrows of development:
Abilities/knowledge Communities of practice Language contact

Individual level interaction Societal level
Figure 5.1 The mutual approximation of the two terms multi-competence and
multilingualism
used in this book, while the term multilingualism is applied all the way to
the individual level, to psycholinguistic and neurolinguistic areas.
This development can be beneficial in at least three different ways.
– MC covers variability on an individual level and gives such behaviour a

status in its own right.
– MC is primarily applied in a psycholinguistic area and now embraces a
wider, interactive, social level in the last dimension. However, it retains
the origins of the term (a strongly individualized view).
– Multilingualism comprises significant sociocultural aspects, in which
all individuals with their embedded language acquisition in a broader or
narrower variety take part.
Elsewhere (Franceschini 2011a) I proposed using the term ‘multilingu-

alism’ as a framework, with MC defining specific aspects of individual
multilingualism that emerge in a social context. With the latest develop-
ments that aim at bridging the gap between cognition, interaction and
social context, the MC paradigm should come to fruition. Regardless of
what future phenomena may be called, the relationship between the three
elements raises many questions that still beg for a more elaborate answer,
for example the question regarding the skill level a multilingual person
needs in order to trigger one of the ‘third qualities’.
5.8 Inner consistency
With the noticeable increase in awareness, the time has come to focus on
the internal structure of multilingualism, and to overcome the former
process that primarily built on a comparison with monolinguals (a group
that is, by the way, not very easily defined either). Instead, it may be best to
assume that a very large number of people are, in various ways, multi-
lingual and able to deal with varieties in a multi-competent way. In
essence, I am arguing in favor of a continuum rather than a clear distinc-
tion between monolingual and multilingual individuals – a view that is
becoming increasingly accepted. The existence of contradictory results

(even in neurobiological research) highlights the need to find appropriate

categories in order to explain the various phenomena found in people
traditionally labelled as ‘multilingual’. It seems as if data gathering were
based, perhaps for simplicity, on the comparison of two extremes (without
taking due account of MC), and researchers may therefore have drawn
conclusions that do not fully reflect reality. It will take some time before
the concept of MC can satisfy the needs of quantitative research.10 And
until then, we rely on currently available data, and interpret them with
due caution.
But let’s venture a look ahead: we need to identify the factors concerning
multilingualism in individuals (adults as well as children) who use several
varieties in their everyday lives; in other words, we need to find the inner
consistency of multilingualism. This includes the definition of several
‘types’ of multilingualism. As mentioned above (see Chapter 3 above), a
former neurobiological study conducted by the research group
Multilingualbrain focused on quantitative narrative language biographies
in order to identify and distinguish the various types of multilingualism
with respect to growing up with several languages. And this approach has
proved very useful.
Based on this multilingual perspective, we then started to explore the
cognitive advantages offered by exposure to different varieties. The fact
that multilingual individuals have cognitive advantages over monolin-
guals has been known for some time (Bialystok 2007; Costa et al. 2008).
These advantages lie in decision-making, control and attention functions
(Abutalebi et al. 2012), all of which are very useful for practical, everyday
life. These advantages have been proved in various studies using a vali-
dated test (the so-called Flanker test or ANT, Attentional Network Test,
which measures attention and executive functions)11 always contrasting
individuals (children as well as adults) that were categorized as either
monolingual or bilingual.
Many researchers wish to measure the performances of bilingual, trilin-
gual and even multilingual children in order to investigate the connections
within a multilingual repertoire. It is not easy to find such children, espe-
cially in the English-speaking world. Europe is in a more fortunate position.
And especially the region of Trentino–Alto Adige/Südtirol in north-east Italy,
just south of the Alpine Brenner Pass, is an area that has a long tradition of
social and individual multilingualism. In some valleys of eastern Südtirol, a
language minority speaks Ladin.12 There, schools integrate compulsorily the
local Ladin varieties, with classes being taught in Ladin, German and Italian.
Furthermore, English is taught from fourth grade.
We have performed the Flanker test with 118 fifth-graders (Videsott et al.
2012).13 For each child, a brief language biography and all school marks, as
well as the teacher’s assessment, were available. While a detailed descrip-
tion of the procedure would go beyond the scope of this contribution (see
Videsott et al. 2012 for more details), I would like to highlight a detail that

might give us important clues with regard to MC. It is important to note that
the Flanker test measures cognitive, not linguistic, capacities: it measures
networks in charge of alerting, orientation and conflict. Previous Flanker tests
(Bialystok 1999, and also Bialystok 2011) established that bilingual children
show more rapid response times and make fewer mistakes than monolin-
gual children. Yet the underlying reasons were still unknown, as finding
differences within a defined multilingual group would require sufficiently
large and socially homogeneous groups of people who grew up with several
languages. In our sample, we correlated the Flanker test results (according to
reaction times and accuracy) with all school marks and classified them
according to their language competence in the four varieties (Ladin,
German, Italian and English). Among all possible correlations – for example
music, mathematics and sports – only one result was significant, and highly
significant at that: the best reaction times and fewest mistakes were con-
sistently made by children that had the highest language scores. This shows
that high language competence correlates with superior cognitive skills.
Children with lower language scores appear to benefit less from the cogni-
tive skills that the test measures.
In other words, an individual that excels in dealing with varieties has the
greatest cognitive advantage: apparently, the use of several languages
from a young age improves cognitive skills. After all, children who grow
up in such contexts are used to taking quick decisions regarding how to
talk, which expression to use, which communication partner speaks
which language and how to adapt to different communication situations.
Taking multiple decisions as to how to react linguistically is an inherent
attribute of these children, and the better their language skills, the more
likely they are to outperform in this task. In a further step – a longitudinal
study of children from South Tyrol and fMRI – we explored which net-
works in the brain are active in this process in order to establish ‘how
neural regions work together to achieve skilled performance’ (Green
2014): an area of bilingualism research that a growing number of research
groups are taking an interest in (see the recent special issue of International
Journal of Bilingualism edited by Reiterer and Festmann 2014). Once again,
we focused on networks that were specifically developed in children who
grow up in multilingual contexts. In other words, this research did not
focus on a comparison between monolingual and bilingual children, but
on the development of a group of multilingual ten-year-old children (see
Della Rosa et al. 2013). Comparing two measurements that were made 18
months apart, we identified an increase in grey matter in a specific area of
the brain, namely the LIPL (left inferior parietal lobe); hence, this area is
part of a network that develops particularly strongly through the experi-
ence of multilingualism. This finding confirmed the hypothesis that ‘the
LIPL may represent the site for dynamic interchange between multilin-
gualism and the neural development of this neurocognitive advantage’
(Della Rosa et al. 2013, p. 605).

We conclude the following:
These results indicate that multilingualism is associated with structural

adaptations in a brain region ruled by a more general structure–function
principle. We suggest that this LIPL structural plasticity traced in early
childhood is the result of dynamic functional requirements (i.e. attention,
memory and phonological categorization functions) that are necessary for
developing high levels of MC [Multilingual Competence] with the effect of
enhancing attention functions deployed in order to monitor and control
the different languages being spoken. (Della Rosa et al. 2013, p. 607)
From here, it is only a small step to concluding that these children have a
high flexibility in their multiple-variety repertoire (this flexibility can be
observed rather well; for more information on written competences see
Franceschini 2013) and that they are, potentially, multi-competent.
The cognitive advantages that can be measured in this as well as in other
studies not only concern the specific networks that are enhanced through
multilingual experience, but also – and this is what makes the described
study so unique – create a link to language competence: a higher level of
‘mastery’ enhances the cognitive advantages. These findings can be asso-
ciated with studies that focus on the identification of contexts in which L2
affects L1 (Cook 2003). Kecskes and Papp (2003), for example, explore the
conceptual effect of L2 on L1. The authors highlight the role of proficiency
after a certain threshold is reached.
Many more questions will have to be addressed in the future, and the
internal consistency of competences that resulted from a lifetime spent with
highly variable linguistic experience needs to be analysed in greater detail.
5.9 Conclusion
The ongoing discussion regarding MC has enlivened multilingualism

research and has prompted a shift in linguists’ perception. Today, MC no
longer merely represents a pluralistic approach to language acquisition
but increasingly embraces the cultural dimension of language acquisition
and use. This is far less trivial than it may seem at first glance, considering
its originally narrow definition, anchored originally in a generativist view-
point, which later transitioned into a psycholinguistic approach initially
focusing exclusively on the individual’s abilities.
Over the course of time, researchers became increasingly aware that the
L2 user uses his or her varieties not simply in a mechanical, instrumental
way: the acquisition and the use of other varieties add a deeper layer to
how the speaker conceptualizes the world they live in. And that is not all:
the speaker’s L1 knowledge can be influenced by the L2.
This conceptualization is not solely due to the individual’s specific
‘cognitive attributes’, but rather due to the fact that the speaker is, with

his or her linguistic abilities, exposed to different cultures and is thus able
to build a deeper understanding thereof (see Alptekin 2010, p. 106).
In this sense and amid a larger European project (i.e. the EU LINEE –
Languages in a Network of European Excellence), we defined MC both as a
tool and as a specific state of mind in which an individual faces all of his or
her life experiences.
Definition of multi-competence in the LINEE project:

Multi-competence, i.e., the knowledge of more than one language in the
mind, is part of the individual capacity of a person and develops in inter-
action with his or her social or educational environment. Multi-competent
individuals make use of their linguistic knowledge when interacting
within a range of linguistic settings, including both multilingual and
monolingual situations. Multi-competence, or multilingual competence,
is thus at the same time a tool and a state and relates to the complex,
flexible, integrative, and adaptable behavior which multilingual indivi-
duals display. A multi-competent person is therefore an individual with
knowledge of an extended and integrated linguistic repertoire who is able
to use the appropriate linguistic variety for the appropriate occasion.
(see Franceschini 2011a, p. 351, with further sources)
This sets a sociolinguistic background for the term MC – a view, in fact,

that emerged organically throughout the four years of our linguistic
projects in various multilingual regions in Europe, in many groups and
communities of practice.
Variable as linguistic behaviour may be, we need to keep in mind that
the standardization process as applied in institutions such as schools has a
direct impact on the variable use of languages. The speaker is not only
variable in his or her language repertoire for the pure joy of it, but rather
because he or she knows from past interaction how to apply all the tools at
his or her disposal and how the use of more or less legitimized varieties
affects certain interaction partners or groups in specific situations (think,
for example, of flirting). Knowing how to adapt linguistically, for example
when talking to teachers or in a confrontation with the police, in order to
avoid sanctions or marginalization, is also considered a tool.
A look at the range covered by MC and multilingualism reveals a process
of convergence between the two terms. The two topics overlap in the area
of supra-individual, multilingual social practice.
This approximation is not new: for quite some time now, an increasingly
psycholinguistic view is combined with the social embedding of linguistic
action and knowledge, with social practice becoming tangible evidence of
individual behaviour: the connection between internal structures of
linguistic phenomena, cognitive processes and social experience is an
urgent issue far beyond SLA (see also Hall et al. 2006, p. 229, and
Kramsch 2009). The two separate approaches – psycholinguistic on one
side, social on the other – are giving way to a more integrated view that

embraces interactive and cultural action (Ortega 2011; Ortega et al. 2013).
This new view does not look for dichotomies where none exist, but instead
focuses on continua, on areas where the phenomena merge into one
another, as in language gliding from a dialect to forms that are closer to
a standardized language version.
This more integrated view (or dare we call it a paradigm?) – long
dismissed as ‘holistic’ and ‘excessively complex’ – is becoming
increasingly formalized outside SLA, and the approaches are beginning
to be cross-referenced: among others, dynamic theories (Herdina and
Jessner 2002; de Bot, Lowie and Verspoor 2007); complexity theories
(Franceschini 2003b; Larsen-Freeman and Cameron 2008; Ortega et al.
2013) or usage-based theories (Tomasello 2005) are being debated as a
means of estimating their validity. It will be interesting to see the future
contributions of multilingualism research to this debate.
Under these circumstances, MC as a term must be seen as a ‘school of
thought’ rather than a final theory: today, MC is an approach that inspires
researchers and that could, considering the impressive amount of research
in the field, become an independent model. We can already see the
outlines of such a model, but it is far from being sufficiently explicit to
be applied in research paradigms in which (quantitative) verification is
needed. To reach the required level of explicitness, further dimensions of
MC should be defined: categories, subcategories, areas . . . Nonetheless,
the term MC is important because – most likely owing to its narrow
definition – it has acted as a catalyst for one of the most important reforms
in the field of SLA over the last decades.
The latest definition of MC (‘the overall system of a mind or a commu-
nity that uses more than one language’) further highlights this increased
openness and an explicit link to the social embedding. However, there is
also a danger of the term becoming too broad and thus losing its distinctive
character – a risk that the term multilingualism also faces. The definition
of subdimensions such as the above-mentioned distinctions in social or
institutional multilingualism as well as discursive and individual multi-
lingualism aims at circumventing these dangers. From today’s venture
point, the term multilingualism stands a chance of remaining the most
comprehensive ‘umbrella term’ (that embraces L2, L3 and Ln). At the same
time, it is becoming increasingly inadequate because it seems to remain
anchored to an additive view on the diversity of languages, especially since
individual speakers do not apply the same distinctions as linguists (e.g. in
multiple language usage). In these cases, the term multilingualism fails to
live up to an emic view, while MC is more neutral and explicitly tries to
convey the transversal processes involved, e.g. between languages.
Ultimately, a more radical view of MC will put an end to interpreting
single languages as autonomous systems. This goes hand in hand with the
need for new concepts to describe emerging phenomena – concepts that
accommodate the dynamics and variability seen in language use. Or, as

Hall et al. (2006, p. 232) rightly pointed out: there is a need for concepts
that are as ‘malleable and porous as the social actions’ they describe.
The inevitable conclusion is that MC can also be applied to individuals
formerly described as monolingual, as their language use can also be
variable, adaptive, flexible and adjusted to individual situations. When
language is seen as a social practice, there is nothing standing in the way
of this view. A speaker that is sufficiently variable in his or her social
practice and his or her affordance thereof can be just as multi-competent
as others. Thus, the ‘number’ of varieties spoken matters very little: a
‘monolingual’ speaker (again, the term is very unsatisfactory!) can be just
as multi-competent within his or her variable repertoire.
Notes
1. I am taking the liberty to abbreviate the name dropping: Selinker

(1972), Lüdi and Py (1984), Gal (1987), Rampton (1995), Auer
(1999), Muysken (2000), Hornberger (2003), Jørgensen (2008),
Creese and Blackledge (2010), Blommaert (2013), and Garcia and
Li Wei (2014).
2. Enriched with student projects. The project ‘Leben mit mehreren
Sprachen: Sprachbiographien tschechischer Remigranten und mehr-
sprachiger Schweizer’ (‘Living with various languages: language bio-
graphies of Czech remigrants and multilingual Swiss’) was
conducted under the programme ‘Cooperation in Science Research
with Central and Eastern European Countries and Newly
Independent States’ (1996–1998) (Swiss National Science
Foundation no. 7CZPJ048495). We labelled it the BPP Project (Basel
Prague Project). The Multilingualbrain project ‘Neurolinguistische
Korrelate der Mehrsprachigkeit im natürlich mehrsprachigen
Umfeld der Regio Basiliensis’ (‘Neuro-linguistic correlates of multi-
lingualism in the naturally multilingual environment of Regio
Basiliensis’) was financed by the University of Basel from 2000 to
2003 (Innovation Fund, decision 00.05.101/08.05.00); C. Nitsch (direc-
tor), R. Franceschini, G. Lüdi, E.-W. Radü. There is still considerable
interest in this area (for more recent studies, see www.languagestu
dies.unibz.it). Regarding language biographies, especially the above-
mentioned projects, see Franceschini (2001a, 2001b, 2001c, 2002,
2003a, 2010), Franceschini and Miecznikowski (2004), Fünfschilling
(1998), Miecnikowski (2001, 2010) and Veronesi (2008a, 2008b,
2008c, 2009, 2010).
3. Even in a generativist view, we have to admit that the language
acquisition device and its effect depend on at least minimal input,
mainly lexical. The positions regarding minimal input and the
consideration of social interaction, of course, diverge.

4. These last-mentioned phenomena are part of linguistic landscape

research: Landry and Bourhis (1997), Backhaus (2007), Shohamy and
Gorter (2009), Shohamy et al. (2010), Cook (2014).
5. This argument does not aim at simplifying matters and declaring the
speaker as the only valid criterion. Rather, the subject matter is far
more delicate and should be dear to any scientist, i.e. the adequacy of
observation and subsequent description: ideally, a phenomenon –
here a phenomenon embedded in social action – should be described
with instruments that reflect the qualities that these actions have for
the participants. If a natural way of dealing with several languages, for
example, does not reflect the quality of switching from one clearly
defined linguistic entity to another, there is a third quality that needs
to be described. Fundamentally speaking, this heuristic view of new
varieties is also known in the area of creole studies. In the cases
derived from the above-mentioned language biographies and the
behaviour in the Central European region, the focus is more on the
fact that the speaker’s repertoire also includes ‘translingual’ beha-
viour – despite the fact that the same speaker is able, in other contexts,
to speak a legitimized language according to socially accepted norms.
6. For details see Bloch et al. (2009, p. 627, Table 1) and Wattendorf et al.
(2001, 2014). Bloch et al. (2009) analysed 44 trilingual subjects across a
range of languages: L2 was at a B1 level in just one language, otherwise
higher; with L3, only two subjects reached B1, the other scored higher.
Sixteen subjects scored C2 in L2, eleven subjects in L3. In this study,
brain areas showing activation were BA 44, 45, 22, 6, 9, 10, 24, and 32,
especially in the left, but partially in the right hemisphere as well
(Bloch et al. 2009, p. 631).
7. We know that linguistic abilities seem to be sensitive to the age of
exposure to further languages: ‘The understanding of these age effects
is complicated by the fact that the different components of language
have different critical periods. In particular, adults rarely attain a
perfect phonological performance although they have the potential
for perfect lexical and syntactic acquisition of a second language
(Scovel 1988). The studies of Kuhl et al. (2003) and Sebastián-Gallés
(2006) demonstrated that while very young infants are able to perceive
all human-language contrasts, between 6 and 12 months they show a
decline in perception of non-native sounds and simultaneously an
improvement in the perception of native sounds’ (Bloch et al. 2009,
p. 631).
8. In this view, varieties are seen as frozen speech patterns and forms
that resulted from a more or less closely knit network of relationships
that emerged from the linguistic habitus of groups. Over time, such
habits crystallized in statements and speech patterns that are clearly
recognizable and have a social and cultural valence. Group varieties
are such examples, as are learner varieties (which gave rise to the term

MC). The designation as a ‘language’ then refers to legitimized forms

that a society has recognized as a norm and that are historically
established through a set of various standardization activities (institu-
tions such as schools, essentially political power constellations, as
seen in (a)).
9. Berthele (2012) describes similar processes as ‘inferability of features’.
10. The fruitful term ‘bilingual mode’ (Grosjean 2001) shares a similar
fate: it is illuminating, but has not yet been fully validated to allow
for an application in quantitative psycholinguistic research. I highly
recommend reading the illuminating criticism by François Grosjean,
who created the term and the paradigm, in Grosjean (1998) and
Grosjean et al. (2003).
11. For the ANT (Attentional Network Test) we used the version of the task
by Costa et al. (2008): ‘In this test participants are asked to indicate the
direction of the central target arrow (left: ←; right: →) flanked by four
other arrows or four horizontal lines. The stimuli appeared either
above or below a fixation point. The test is characterised by two
factors: Cue Type and Flanker Type. The latter refers to the position of
the arrows in the space and is composed of three conditions: Congruent,
when the arrows are all pointing in the same direction; Incongruent,
when the central arrow is pointing to the opposite direction than the
other four arrows; and Neutral, when the central arrow is accompanied
by four horizontal lines (a). The Cue Type, instead, refers to the pre-
sentation of different combinations of asterisks before the appearance
of the arrows. That factor is composed of four levels: the No Cue, the
basic level, characterised by the failure of any asterisk, the Central Cue,
that consists of the appearance of an asterisk in the center of the
screen; the Double Cue, i.e. the presentation of two asterisks (one
above and one below the fixation point); and the Spatial Cue, an asterisk
that predicts the subsequent position of the arrows (b). The crossing of
these two factors determines twelve experimental conditions for a
total amount of 96 trials. B. Schematic representation of the experi-
mental procedure with duration time of the single stimuli. The sample
shows the Incongruent Condition preceded by a Spatial Cue. The
single stimuli are presented in the following order: it starts with a
fixation cross that appears in the centre of the computer screen. After
400 ms one of the four cue types compares for only 100 ms (except for
the No Cue). The fixation cross appears again (400 ms) followed by the
arrows, which remain on the screen for max. 1700 ms’ (Videsott et al.
2012, p. 887).
12. A Romance variety genetically related to neighbouring varieties of
Rhaeto-Romance (in Switzerland) and Friulian (in north-eastern
Italy). In the Autonomous Province of Bolzano, approximately 20,000
people state that they speak Ladin (see ASTAT: http://www.provincia
.bz.it/astat/de/default.asp), mainly in the valleys of Gardena and Badia.

It is also spoken in the valleys of Fassa and Fodom. While its use is
slowly declining, it is actively used in a family context.
13. The study was conducted in 2008–2012 by the University of Bolzen–
Bolzano’s Language Study Unit, in cooperation with the Università
Vita–Salute San Raffaele. It was supported by funds of the Research
Committee of the Free University of Bozen–Bolzano (www.languages
tudies.unibz.it).
References
Abutalebi, J., Della Rosa, P. A., Green, D. W., Hernandez, M., Scifo, P.,
Keim, R., Cappa, S. F. and Costa, A. 2012. ‘Bilingualism tunes the
anterior cingulate cortex for conflict monitoring’, Cerebral Cortex 22,
2076–2086.
Adams, J. N., Janse, M. and Swain, S. (eds) 2002. Bilingualism in Ancient Society:
Language Contact and the Written Text. Oxford University Press.
Alptekin, C. 2010. ‘Redefining multi-competence for bilingualism and
ELF’, International Journal of Applied Linguistics 20, 95–110.
Auer, P. (ed.) 1998. Code-switching in Conversation: Language, Interaction and
Identity. London: Routledge.
Auer, P. 1999. ‘From codeswitching via language mixing to fused lects:
toward a dynamic typology of bilingual speech’, International Journal of
Backhaus, P. 2007. Linguistic Landscapes: Comparative Study of Urban
Multilingualism in Tokyo. Clevedon: Multilingual Matters.
Berthele, R. 2012. ‘Multiple languages and multiple methods: Qualitative
and quantitative ways of tapping into the multilingual repertoire’, in E.
Berthele (ed.), Methods in Contemporary Linguistics: In Honour of Iwar Werlen,
pp. 195–218. Berlin: De Gruyter Mouton.
Bialystok, E. 1999. ‘Cognitive complexity and attentional control in the
bilingual mind’, Child Development 70, 636–644.
Bialystok, E. 2007. ‘Cognitive effects of bilingualism: how linguistic experi-
ence leads to cognitive change’, International Journal of Bilingual Education
and Bilingualism 10, 210–223.
Bialystok, E. 2011. ‘Coordination of executive functions in monolingual
and bilingual children’, Journal of Experimental Child Psychology 110,
461–468.
Bloch, C., Kaiser, A., Kuenzli, E., Zappatore, D., Haller, S., Franceschini, R.,
Luedi, G., Radue, E.-W. and Nitsch, C. 2009. ‘The age of second language
acquisition determines the variability in activation elicited by narration
in three languages in Broca’s and Wernicke’s area’, Neuropsychologia
47 (3), 625–633.
Blommaert, J. 2013. Ethnography, Superdiversity and Linguistic Landscapes.

Cook, V. J. 1991. ‘The poverty-of-the-stimulus argument and multi-compe-

tence’, Second Language Research 7, 103–117.
Cook, V. J. 1992. ‘Evidence for multi-competence’, Language Learning 42,
557–591.
Cook, V. J. (ed.) 2003. Effects of the Second Language on the First. Clevedon:
Cook, V. J. 2007. ‘The nature of the L2 user’, in L. Roberts, A. Gurel, S. Tatar
and L. Marti (eds), EUROSLA Yearbook 7, 205–220. Reprinted in Li Wei (ed.)
2011, The Routledge Applied Linguistics Reader, pp. 77–89. Routledge.
Cook, V. J. 2014. ‘Meaning and material in the language of the street’, Social
Semiotics 25, 181–109.
Costa, A., Hernandez, M. and Sebastián-Gallés, N. 2008. ‘Bilingualism aids
conflict resolution: evidence from the ANT task’, Cognition 106, 59–86.
Creese, A. and Blackledge, A. 2010. ‘Translanguaging in the bilingual class-
room: a pedagogy for learning and teaching?’, Modern Language Journal
94, 103–115.
De Bot, K., Lowie, W. and Verspoor, M. 2007. ‘A Dynamic Systems Theory
approach to second language acquisition’, Bilingualism: Language and
Cognition 10 (7–21), 51–55.
Della Rosa, P. A., Videsott, G., Borsa, V., Canini, M., Weekes, B. S.,
Franceschini, R. and Abutalebi, J. 2013. ‘A neuronal interactive location
for multilingual talent’, Cortex 49, 605–608.
Dirim, I. and Auer, P. 2004. Türkisch sprechen nicht nur die Türken: Über die
Unschärfebeziehungen zwischen Sprache und Ethnie in Deutschland [Turkish is
not only spoken by Turks: on blurred boundaries between language and
ethnicity in Germany]. Berlin: De Gruyter.
Franceschini, R. 1998. ‘Code-switching and the notion of code in linguis-
tics: proposals for a dual focus model’, in P. Auer (ed.), pp. 51–74.
Franceschini, R. 1999. Italiano di Contatto: Parlanti Occasionali e Riattivazioni di
Conoscenze Non Focalizzate [Contact Italian: occasional speakers and unfo-
cused knowledge]. Habilitation. Universität Basel.
Franceschini, R. 2001a. ‘Sprachbiographien randständiger Sprecher’
[Linguistic biographies of marginal speakers], in R. Franceschini (ed.),
pp. 111–125.
Franceschini, R. 2001b. ‘Der “Ko-Adiuvat”: die Figur der Stützperson im
sprachbiographischen Interview mehrsprachiger Sprecher’ [The ‘co-
helper’: the support figure in linguistic biographies], in T. Keller and F.
Raphaël (eds), pp. 227–238.
Franceschini, R. (ed.) 2001c. Biographie und Interkulturalität: Diskurs und
Lebenspraxis [Biography and interculturality: discourse and life practices].
Tübingen: Stauffenburg-Verlag.
Franceschini, R. 2002. ‘Sprachbiographien: Erzählungen über Mehr-
sprachigkeit und deren Erkenntnisinteresse für die Spracherwerbs-
forschung und die Neurobiologie der Mehrsprachigkeit’ [Linguistic
biographies: narrations about multilingualism and their interest for

language acquisition and the neurobiology of multilingualism], in K.

Adamzik and E. Roos (eds), Biografie Linguistiche / Biographies Langagières /
Biografı́as Lingüı́sticas / Sprachbiografien, pp. 19–33. Neuchâtel: Université
de Neuchâtel, Bulletin VALS-ASLA, 76.
Franceschini, R. 2003a. ‘Unfocussed language acquisition? The presenta-
tion of linguistic situations in biographical narration’, Forum Qualitative
Sozialforschung / Forum Qualitative Social Research 4, art. 19.
Franceschini, R. 2003b. ‘Modellbildung über die Mehrsprachigkeit hinaus:
für eine Linguistik der Potentialität (LP)’ [Model building beyond multi-
lingualism: for a linguistics of potentiality], in L. Mondada and S.
Pekarek Doehler (eds), Plurilinguisme / Mehrsprachigkeit / Plurilingualism:
Enjeux Identitaires, Socio-culturels et éducatifs, pp. 247–259. Tübingen:
Francke.
Franceschini, R. 2009. ‘The genesis and development of research in multi-
lingualism: perspectives for future research’, in L. Aronin and B.
Hufeisen (eds), The Exploration of Multilingualism: Development of Research
on L3, Multilingualism and Multiple Language Acquisition, pp. 27–61.
Amsterdam: Benjamins.
Franceschini, R. (ed.) 2010. Sprache und Biographie [Language and biography]:
Zeitschrift für Literaturwissenschaft und Linguistik (LiLi), vol. 40.
Franceschini, R. 2011a. ‘Multilingualism and multi-competence: a concep-
Franceschini, R. 2011b. ‘What unfocussed acquisition can say about rela-
tive openness of systems in contact’, in P. Ramat and E. Miola (eds),
Language Contact and Decay: Socio-political and Linguistic Perspectives,
pp. 109–135. Pavia: IUSS.
Franceschini, R. 2012. ‘Unfokussierter Spracherwerb in Kontaktsituationen:
Sprachexposition als Teil des Sprachwandels’ [Unfocused language acqui-
sition in contact situations: exposition as part of language change],
Sociolinguistica 26, 41–57.
Franceschini, R. 2013. ‘Die Entwicklung dreisprachiger Schreib-
kompetenzen: Resultate aus den ladinischsprachigen Tälern
Südtirols’ [The development of trilingual literacy competencies:
results from the Ladin-speaking valleys of South Tyrol], in J. Erfurt,
T. Leichsering and R. Streb (eds), Mehrsprachigkeit und Mehrschriftigkeit:
Sprachliches Handeln in der Schule, pp. 57–78, vol. 83 of Osnabrücker
Beiträge zur Sprachtheorie (OBST). Duisburg: Univ.-Verl. Rhein–
Ruhr.
Franceschini, R. and Miecznikowski, J. (eds) 2004. Leben mit Mehreren
Sprachen: Sprachbiographien im Mitteleuropäischen Kontext [Living with
several languages: linguistic biographies in Mitteleuropa]. Bern: Lang.
Fünfschilling, J. 1998. ‘Spracherwerb als Teil der Biographie: zur
Versprachlichung von Erwerbserinnerungen in narrativen Interviews’
[Language acquisition as part of one’s biography], in L. Mondada and
G. Lüdi (eds), Dialogue entre Linguistes: Recherches en Linguistique à l’Institut des

Langues et Littératures Romanes de l’Université de Bâle. Acta Romanica

Basiliensia (ARBA), vol. 8, pp. 65–79. Basel Universität.
Gal, S. 1987. ‘Linguistic repertoire’, in U. Ammon, N. Dittmar and
K. Mattheier (eds), Sociolinguistics: An International Handbook of the Science
of Language and Society / Soziolinguistik: Ein Internationales Handbuch zur
Wissenschaft von Sprache und Gesellschaft, vol. 1, pp. 286–292. Berlin, New
York: De Gruyter.
Education. New York: Palgrave.
Gardner-Chloros, P. 1991. Language Selection and Switching in Strasbourg.
Oxford: Clarendon Press.
Green, D. W. 2014. ‘Foreword’, International Journal of Bilingualism 18 (1), 5.
Grosjean, F. 1998. ‘Studying bilinguals: methodological and conceptual
issues’, Bilingualism: Language and Cognition 1, 131–149.
Grosjean, F. 2001. ‘The bilingual’s language modes’, in J. L. Nicol (ed.), One
Mind, Two Languages, pp. 1–22. Oxford: Blackwell.
Grosjean, F., Li, P. M. T. and Rodriguez-Fornells, A. 2003. ‘Imaging bilin-
guals: when the neurosciences meet the language sciences’, Bilingualism:
Hall, J. K., Cheng A. and Carlson M. T. 2006. ‘Reconceptualizing multi-
competence as a theory of language knowledge’, Applied Linguistics 27,
220–240.
Herdina, P. and Jessner, U. 2002. A Dynamic Model of Multilingualism:
Perspectives of Change in Psycholinguistics. Clevedon: Multilingual Matters.
Hornberger, N. (ed.) 2003. Continua of Biliteracy: An Ecological Framework for
Educational Policy, Research and Practice in Multilingual Settings. Clevedon:
Jørgensen, J. N. 2008. ‘Polylingual languaging around and among children
and adolescents: introduction’, International Journal of Multilingualism 5 (3),
161–176.
Kecskes, I. 1998. ‘The state of L1 knowledge in foreign language learners’,
Word 49, 321–340.
Kecskes, I. and Papp, T. 2003. ‘How to demonstrate the conceptual effect of
L2 on L1?’, in V. J. Cook (ed.), pp. 247–265.
Keller, T. and Lüdi, G. (eds) 2008. Biographien und Staatlichkeit / Biographies et
Pratiques de l’état [Biographies and governmental practices]. Berlin:
Berliner Wissenschaftsverlag.
Keller, T. and Raphaël, F. (eds) 2001. Biographies au Pluriel / Biographien im Plural
[Biographies in plural]. Strasbourg: Presses Universitaires de Strasbourg.
Kim, K. H. S., Relkin, N. R., Lee, K. L. and Hirsch, J. 1997. ‘Distinct cortical
areas associated with native and second languages’, Nature 388, 171–174.
Kloss, H. 1967. ‘Abstand-languages and Ausbau-languages’, Anthropological
Linguistics 9, 21–41.
Kramsch, C. 2009. The Multilingual Subject: What Foreign Language Learners Say
about their Experience and Why it Matters. Oxford University Press.

Kuhl, P. K., Tsao, F. M. and Liu, H. M. 2003. ‘Foreign-language experience in

infancy: effects of short-term exposure and social interaction on pho-
netic learning’, Proceedings of the National Academy of Sciences of the United
States of America 100, 9096–9011.
Landry, R. and Bourhis, R. 1997. ‘Linguistic landscape and ethnolinguistic
vitality: an empirical study’, Journal of Language and Social Psychology 16,
23–49.
Larsen-Freeman, D. and Cameron, L. 2008. Complex Systems and Applied
Linguistics. Oxford University Press.
Lévi-Strauss, C. 1962. La Pensée Sauvage. Paris: Plon. Translated as The
Savage Mind, The University of Chicago Press, 1966.
Lüdi, G. and Py, B. 1984. Zweisprachig durch Migration: Einführung in die
Erforschung der Mehrsprachigkeit am Beispiel zweier Zuwanderergruppen in
Neuenburg (Schweiz) [Bilingual by migration: introduction to the study of
multilingualism with examples of two groups of migrants in Neuchâtel
(Switzerland)]. Tübingen: Niemeyer.
Miecznikowski, J. 2001. ‘Parallelen in der Darstellung von Sprachkontakt
und -erwerb: zwei biographische Interviews’ [Parallels in the presenta-
tion of language contact and acquisition], in T. Keller and F. Raphaël
(eds), pp. 215–226.
Miecznikowski, J. 2010. ‘Mehrsprachig leben und studieren: die Sicht
studentischer Migranten’ [Multilingual biographies and academic envir-
onments: student migrants’ accounts], Sprache und Biographie: Zeitschrift
für Literaturwissenschaft und Linguistik (LiLi) 40, 129–150.
Muysken, P. 2000. Bilingual Speech: A Typology of Code-mixing. Cambridge
University Press.
Myers-Scotton, C. 1993. Duelling Languages: Grammatical Structure in
Codeswitching. Oxford: Clarendon Press.
Ortega, L. 2011. ‘SLA after the social turn: where cognitivism and its
alternatives stand’, in D. Atkinson (ed.), Alternative Approaches to Second
Language Acquisition, pp. 167–180. London: Routledge.
Ortega, L., Cumming, A. and Ellis, N. C. (eds) 2013. Agendas for Language
Learning Research, vol. 63 of Currents in Language Learning. Wiley-Blackwell.
Rampton, B. 1995. Crossing: Language and Ethnicity among Adolescents. London:
Longman.
Reiterer, S. and Festmann, J. 2014. Multilingual Brains: Individual Differences
in Bi- and Multilinguals, special issue of International Journal of
Bilingualism 18.
Schütze, F. 1987. Das Narrative Interview in Interaktionsfeldstudien [The narra-
tive interview in interactional field work]. Hagen: Gesamthochschule
Hagen.
Scovel, T. 1988. A Time to Speak. New York: Newbury House.
Sebastián-Gallés, N. 2006. ‘Native-language sensitivities: evolution in the
first year of life’. Trends in Cognitive Sciences 10, 239–241.
Selinker, L. 1972. ‘Interlanguage’, IRAL 10, 209–231.

Shohamy, E., Ben-Rafael, E. and Barni, M. (eds) 2010. Linguistic Landscape in

the City. Clevedon: Multilingual Matters.
Shohamy, E. and Gorter, D. (eds) 2009. Linguistic Landscape: Expanding the
Scenery. New York: Routledge.
Tomasello, M. 2005. Constructing a Language: A Usage-based Theory of Language
Acquisition. Cambridge, MA: Harvard University Press.
Veronesi, D. 2008a. ‘Languages and language learning in South Tyrol: some
data from Italian speakers’ language biographies’, International Journal of
Anthropology 23, 261–274.
Veronesi, D. 2008b. ‘Rappresentazioni di lingue e di percorsi di apprendi-
mento: la metafora nelle biografie linguistiche di parlanti in Alto Adige–
Südtirol’ [Representations of languages and language acquisition
pathways: the metaphor in linguistic biographies of speakers in South
Tyrol], in C. Casadio (ed.), Vie della Metafora: Linguistica, Epistemologia,
Psicologia, pp. 120–145. Corfinio: PrimeVie.
Veronesi, D. 2008c. ‘Geschichte, Sprachpolitik und Lebenserzählungen:
erste Gedanken zu Sprachbiographien in Südtirol’ [History, language
policy and live history: first thoughts on linguistic biographies in
South Tyrol], in T. Keller and G. Lüdi (eds), pp. 123–154.
Veronesi, D. 2009. ‘ “Due lingue che sono entrambe mie”: biografie di
parlanti “bilingui” e “monolingui” in un territorio di confine’ [‘Two
languages which are both mine’: biographies of ‘bilinguals’ and ‘mono-
linguals’ in a border country], in I. Riccioni (ed.), Multiculturalismi a
Confronto, pp. 39–56. Bozen–Bolzano University Press.
Veronesi, D. 2010. ‘ “Zu wem ghör i jetz?” bzw. “due lingue che sono
entrambe mie”: sprachbiografien ein- und zweisprachiger Sprecher
aus einem Grenzgebiet’ [‘To whom do I belong now?’ viz. ‘Two languages
which are both mine’: linguistic biographies of mono- and bilingual
speakers from a border country], Sprache und Biographie: Zeitschrift für
Literaturwissenschaft und Linguistik (LiLi) 40, 83–106.
Videsott, G., Della Rosa, P. A., Wiater, W., Franceschini, R. and Abutalebi, J.
2012. ‘How does linguistic competence enhance cognitive functions in
children? A study in multilingual children with different linguistic com-
petences’, Bilingualism: Language and Cognition 15 (4), 884–895.
Wattendorf, E., Festmann, J., Westermann, B., Keil, U., Zappatore, D.,
Franceschini, R., Lüdi, G., Radü, E.-W., Münte, T. F., Rager, G. and
Nitsch, C. 2014. ‘Early bilingualism influences early and subsequently
later acquired languages in cortical regions representing control func-
tions’, International Journal of Bilingualism 18, 57–77.
Wattendorf, E., Westermann, B., Zappatore, D., Franceschini, R., Lüdi, G.,
Radü, E.-W. and Nitsch, C. 2001. ‘Different languages activate different
subfields in Broca’s area’, Neuroimage 13, 624.
Mouton De Gruyter.

6
Multi-competence and
dynamic/complex
systems
Kees de Bot
While the link between multi-competence and Dynamic Systems Theory

(DST)1 is longstanding through the work by Herdina and Jessner (2002),
some issues on the compatibility of the two ways of thinking remain to be
clarified. There are two components in the term multi-competence: ‘multi’
and ‘competence’. Both concepts are not without problems from a DST
perspective. As Cook (Chapter 1, this volume) argues, some aspects of the
concept of multi-competence go very well with basic characteristics of
DST, such as the embeddedness of systems and the constant change. But,
as Cook also indicates, the notion of what constitutes a language is
unclear. To put it bluntly: multi what is multi-competence? Is the assump-
tion that there are multiple but connected entities in the brain that deal
with language in one way or another, and are languages separate sets of
modules in the mind? Cook (Chapter 1, p. 18) takes a clear position in this:
the mental representation of language is a complex system with all sorts
of internal and external relationships; it may be quite arbitrary to divide a
bilingual system into separate areas, modules and subsystems, that can be
called languages in the plural.
First we will discuss some of the core characteristics of DST and indicate
how they relate to multi-competence. Then we will look at the assumption
of languages in the brain and argue that this is unlikely, which has a
significant impact on what multi-competence is.
6.1 Core characteristics of DST
In recent years approaches to language and language development based

on Dynamic Systems Theory (DST) and Complexity Theory have emerged.
An earlier version of this article was presented at a conference in Oxford in June 2009. The author is indebted to Vivian
Cook for his very helpful suggestions for improvement.

126 K E E S D E B OT
A detailed discussion of DST is beyond the scope of the present contribu-

tion, and there are a number of publications that provide more detailed
accounts of language as a dynamic system (de Bot, Lowie and Verspoor
2005, 2007; Larsen-Freeman and Cameron 2008; van Geert 2008, 2011;
Verspoor, Lowie and de Bot 2010).
We start from van Geert’s (1994) definition of dynamic systems: ‘We
define a dynamic system as a set of variables that mutually affect each
other’s changes over time’ (p. 50). So the main components are sets of
variables and change of time. The difference from more static approaches
is that the variables are studied in their interaction over time. The argu-
ment is that different variables (e.g. motivation to learn a language, suc-
cess in learning a language, contact with a language) do not have a fixed
effect, but that they interact and that that interaction itself changes over
time, so not only do motivation and success interact, but this interaction
changes as well. The main characteristics of a dynamic systems based
approach are listed below.
Sensitive dependence on initial conditions. In some systems, small
differences in initial conditions may have a big impact in the long run,
while in other systems large differences in initial conditions have hardly
any effect over time and level out. We have to keep in mind that most
systems have no real beginning and that we are only looking at particular
windows in time: the best-known example of a dynamic system is the
weather system, and that has no real beginning in time. The initial state
consists of all the characteristics of the system, so for second language
development (SLD) that includes proficiency in the first language and
other languages, but also intelligence, working memory capacity, socio-
economic status, social setting, attitudes and so on. For language develop-
ment, this means that small differences between learners may become
large differences and that the same treatment (approach in education)
does not necessarily lead to convergence.
Complete interconnectedness. Dynamic systems are typically
nested: large systems consist of subsystems and these in turn consist of
sub-subsystems and so on. If we focus on language, different subsystems of
the cognitive system play a role: memory, attention and perception. In the
learning process these subsystems, and several others, such as motivation
and language aptitude, contribute. Language may be one of these subsys-
tems, but it could also be viewed as the intersection of a set of subsystems.
All these subsystems are interconnected, which means that change in one
subsystem leads to changes in all other subsystems. For the link with
multi-competence, this is a crucial aspect: because human cognition is
embodied, distributed and shared, there is no clear boundary between the
social and the cognitive. Humans are embedded in their social setting just
as the components of their cognitive system are embedded. Therefore
there is continuous interaction between the social and the cognitive
system.

Multi-competence and dynamic/complex systems 127
Non-linearity in development. This means that the relation

between cause and effect in a developing system is disproportionate.
For language attrition, there may be a disproportionate relation
between time of non-use and decline in the system. In a study on
first language attrition in Dutch migrants in France, de Bot, Gommans
and Rossing (1997) showed that there was no linear relationship
between the length of the period of non-use of the language and the
level of attrition.
Change and development through internal reorganization
and interaction with the environment. The classic example of self-
reorganization is a pile of sand (Goles et al. 2002). When grains of sand fall
on a flat surface, they will build a pile. As grains continue to fall, the pile of
sand will change shape and settle for a while and then change again. The
pile of sand will reorganize itself constantly based on the internal forces in
the system. Why a system reorganizes in a particular way cannot be
predicted because the total of forces and variables cannot be calculated.
Language is also a self-reorganizing system: input will affect the system
that will sometimes simply adopt a new element without internal reorga-
nization, and sometimes adapt itself to the new input in Piaget’s terms, but
there may also be internal forces that lead to reorganization. An example
could be the development of the lexicon. The acquisition of new words
leads to a reorganization of the system: when a learner knows the word
constantly, she will use it in many contexts, but then she finds out that there
are various near-synonyms, like frequently, often, continuously and so on. The
lexical network that constantly is part of will change to accommodate the
new words, giving them each their slightly different meaning, so limiting
the meaning of constantly itself.
– Systems are constantly changing and only temporarily settle into so-called
‘attractor states’. The pile of sand example just described also serves to
show that systems are constantly changing as long as there are forces
working on or in it. Even within a pile of sand there are forces working:
gravity will pull grains down, changes in humidity or temperature will
lead to changes in adhesion between grains, the surface the pile is on will
never be absolutely motionless and so on. Due to these forces the system
will change, but not continually: layers of sand will move and then settle
for some time. When the system will change and what will be the next
attractor state cannot be predicted: it is the result of all the forces inter-
acting over time. In language learning fossilization and stages of develop-
ment can be seen as attractor states. A well-known example is the
pronunciation of the word variable which in standard English has the stress
on the first syllable variable, but which by students and academics from a
wide range of linguistic backgrounds is often pronounced as variable, with
stress on the second syllable. No matter how often they hear the correct
rendering of this word, they will continue using their deeply entrenched
version.

Dependence on internal and external resources. The forces work-

ing in a pile of sand could be seen as resources for change, but in human
learning other types of resources play a role. Internal resources include
motivation to learn, previous knowledge and aptitude. External resources
include opportunities to learn, time invested by the environment, avail-
ability of tools for learning and material conditions (food, shelter) that
make learning possible. The carrying capacity of a system is the sum of the
resources that can be used for optimal development and performance. The
carrying capacity of a highly motivated, talented and well-educated lear-
ner in optimal learning conditions are markedly larger than that of a
poorly motivated, untalented learner with little time and opportunity to
learn. Interesting examples are some of the famous savants speaking many
languages, like Daniel Tammett, who managed to learn to speak Icelandic
fluently within a week. As Erard (2012) mentions in his book on language
savants, their ability to learn a language extremely fast is based on some
very special resources, such as a working memory capacity that is twice
that of the average, less gifted learner, and the ability to focus extremely
on a specific task.
Systems may show chaotic variation over time. Over time a
dynamic system will move from one attractor state to the next, but what
the next attractor state will be is unpredictable. Because of the complex
interaction of variables, the same system with identical initial conditions
may show different attractor states. This unpredictability has been
labelled ‘chaos’ which of course it partially overlaps in meaning with the
more everyday use of the word chaos. The bedroom of a 14-year-old may be
chaos from the parents’ perspective but it is not really unpredictable: in
fact chaos is the normal state in that setting. For language learning, chaos
is not the most likely state: although there is a lot of variation between
individuals, the developmental track in L1 and L2 acquisition also shows
considerable inter-individual overlap. Research by Chan, Lowie and de Bot
(2015) on the developmental paths of writing skills in English by two
Taiwanese twins shows that while their genetic makeup and learning
situation are highly similar, they nonetheless show partial overlap in
their learning, but also considerable differences in terms of lexical and
syntactic complexity. This partial uniformity in development may reflect
either system-internal constraints on development (which is basically
what proponents of the role of Universal Grammar argue) or a constrain-
ing effect of the language input. Smith et al. (2003) point out that the
poverty of input a child receives in language development is in itself the
most important force shaping language: language is the way it is because it
can be acquired on the basis of restricted amounts of input.
Development is conceived of as an iterative process. This means
that the present state of development depends critically on the previous
level of development. Language evolves gradually and every use of the
language is an iterative step in its development. Iteration is the repeated

Table 6.1 Six meanings of ‘language’ (from Cook 2010)
Lang1 a human representation system

Lang2 an abstract external entity
Lang3 a set of actual or potential sentences
Lang4 the possession of a community
Lang5 the knowledge in the mind of an individual
Lang6 a form of action
contact with and use of specific linguistic information, e.g. on past tense
formation. It is not the case that every instance encountered will change
the system, but it will build up till a reorganization takes place. Similarly
language attrition goes in steps in which the present state of the system is
crucial for the next phase.
The main point is that language can be seen as a dynamic system and
that language development is a dynamic process in which the language
system may grow or decline depending on resources and interactions
between input and internal forces. There is no fundamental difference
between the processes of growth and decline: they are both governed by
the same mechanisms. All language users are constantly going through
phases of growth and decline, and their language systems, be it their first
language or any other language, are constantly changing depending on
internal reorganization and input from the environment.
So far, the notion of what is a language has not been addressed, but for a
theory of development, some sort of definition is called for. In his
‘Prolegomena to second language learning’, Cook (2010) distinguishes six
perspectives on language (see also Chapter 1), as shown in Table 6.1.
In my view the six perspectives are not mutually exclusive, and from a
DST perspective language is – at the same time – a set of possible utter-
ances, the possession of a community and the knowledge in the mind of an
individual. And maybe even a form of action. Because language is an
embedded system, both socially and cognitively, the distinction between
the sociolinguistics and the psycholinguistics of language fades. By label-
ling language as just the knowledge in the mind of an individual, the
embeddedness is lost, by defining it as a set of possible utterances the
social embedding is lost. Language resides in the community as much as it
does in the individual brain. The recent work on semantic and syntactic
priming by Hartsuiker and Pickering (2007) shows that the idea of the
completely autonomous language producer that features in models like
that of Levelt (1989) is untenable, since even the psycholinguistic mechan-
isms are influenced by the social setting, but also the utterances produced
by the interlocutor. So when talking about language as a dynamic system
there is no clear distinction between the Lang1–Lang6 perspectives, though
depending on what is focused on, one facet may be more prominent than
another.

A somewhat problematic characteristic of a DST approach is that lan-

guage development cannot really be explained: the factors playing a role
in development may all be known, but how their interaction will change
in the course of development cannot be predicted and may in that sense be
chaotic. A system develops in a certain way because that is how this system
develops. On the one hand this is not a very satisfactory situation, because
traditionally research has been about hypothesis generation and hypoth-
esis testing. The full interconnectedness of systems means that isolating
single variables as explanatory may be unproductive because the impact of
that variable on all the other variables is not taken into account. For the
moment it seems wise to accept that we cannot explain development and
to focus on adequate descriptions of developmental processes.
As this short view on DST shows, it is in many respects compatible with
the premises of multi-competence. The three premises discussed by Cook
(Chapter 1, this volume) can each be linked to aspects of DST.
– Premise 1: concerns the total language system (L1, L2, Ln) in a single
mind or community and their inter-relationships.
The embeddedness and inter-relatedness between subsystems also holds

for the language system. DST also assumes that subsystems interact with
each other over time. The consequences of a DST perspective also imply
that what applies to the individual mind also applies to the community of
users, with the individual embedded in the group. But as mentioned, the
underlying assumption that there are different languages in the mind –
how else could there be inter-relationships? – is debatable, as we will argue
further on.
– Premise 2: multi-competence does not depend on the monolingual

native speaker.
As argued elsewhere (de Bot 2004) and by many other people, monolin-
guals don’t exist if we take styles and registers as systems that are indis-
tinguishable from languages. In DST there is no internal norm: the system
develops through interaction with the environment and internal
reorganization.
– Premise 3: affects the whole mind, i.e. all language and cognitive sys-
tems, rather than language alone.
The formulation of this premise has a distinctively modular flavour. It

suggests that there are different languages and cognitive systems in the
mind that are modular. From a DST perspective, cognition, and therefore
language use, is continuous rather than modular, following the argumen-
tation presented by Spivey (2007). His main argument is that there is
substantial evidence against the existence of separate modules for specific
cognitive activities such as face recognition and object recognition and that
there are no grounds for assuming that things are different for language

use. We will not go into this any further but will focus on two issues that
seem to make the compatibility of DST and multi-competence problematic.
First we will look at what it is that is ‘multi’ in multi-competence.
6.2 Languages as entities
Ever since Weinreich’s influential book Languages in Contact from 1953,

linguists of different hyphenation have looked at languages as entities that
can be studied in the same way as biologists study different species: lan-
guages are separate entities that have their own history and characteristics.
For various sociopolitical reasons, languages have been defined and labelled
and linguists have studied them in various forms looking at ‘the grammar of
French’ or ‘the Spanish morphological system’. A common idea has been
that there is a direct link between what descriptive and theoretical linguists
have to say about language and how language in used in individuals and
groups. In my view there are now good reasons to review these positions and
at least consider alternatives for it. There seem to be arguments from
different fields that call into question both the link between description
and use and the idea that there are separate languages in our head.
In the structuralist tradition, languages were seen as systems of rules
that were consistent and could be described with ever more advanced tools
and paradigms. A language could be defined as a unique set of specific
rules and elements. Languages may share some elements but each lan-
guage is its own fixed and unique set. Speakers of that language are
assumed to be highly similar with largely overlapping sets.
An alternative view on what constitutes a language has been developed
by researchers working from an emergentist perspective:
The term ‘emergentism’ emphasizes the progressive nature of the process
(of language development), namely that qualitatively new and more com-
plex representations can emerge on the basis of knowing its simpler
component parts. (Behrens 2009, p. 40)
The most outspoken representative of the emergentist movement is prob-

ably Paul Hopper. An often quoted statement comes from his 1998 article:
There is no natural fixed structure to language. Rather, speakers borrow
heavily from their previous experiences of communication in similar
circumstances, on similar topics, and with similar interlocutors.
Systematicity, in this view, is an illusion produced by the partial settling
or sedimentation of frequently used forms into temporary subsystems.
(Hopper 1998, pp. 157–158)
What this boils down to is that there are no separate languages but only
discourses. This means that if we accept this position, we need to rethink
what languages are both in terms of development and cognitive

processing, but also what constitutes a language or languages from the

multi-competence perspective.
Hopper argues that the idea that we acquire a language that then repre-
sents a fixed set in our mind is wrong:
This means that the task of ‘learning a language’ must be reconceived.
Learning a language is not a question of acquiring grammatical structure
but of expanding a repertoire of communicative contexts. Consequently,
there is no date or age at which the learning of language can be said to be
complete. New contexts, and new occasions of negotiation of meaning,
occur constantly. A language is not a circumscribed object but a loose
confederation of available and overlapping social experiences … That
what adults know, and what children learn, is not an abstract system of
units with meanings and rules for combining them, but are integrated
normative modes of interactive behavior and the accompanying social use
of corporeal signs such as words and gestures, to which concepts like
language and grammar are almost entirely secondary.
(Hopper 1998, pp. 171–172)
Because Hopper sees language not as an entity, but rather as ‘the set
of sedimented conventions that have been routinized out of the more
frequently occurring ways of saying things’ (p. 161), there is little
reason to assume different languages in bilinguals, which is a rather
radical deviation from mainstream thinking at the moment. In his
writings, Hopper never explicitly refers to the repercussions of his
ideas for bilingualism, though these are far-reaching.2 Recently,
Larsen-Freeman and Cameron (2008) have proposed a link between
Hopper’s emergentism and a dynamic systems perspective on lan-
guage development. The dynamic perspective implies that the lan-
guage system is constantly changing due to external input and
internal restructuring. They therefore argue against the idea of a
language as a fixed and stable entity:
While it is convenient to use a simplifying metaphor and conceive of the
language system contained somehow as a static entity in someone’s head,
we think this is not an apt way to think of a person’s language resources.
(Larsen-Freeman and Cameron 2008, p. 125)
6.2.1 Transfer and code-switching: the assumption of languages

as separate entities
The assumption of languages as separate entities has had a tremendous
impact on research in the area of second language acquisition (SLA). In
the early days extensive research was carried out on Contrastive Analysis
and its predictions, the idea being that contrast between language sys-
tems leads to problems in learning while similarity leads to facilitation.
The predictions of CA have not come out too well, but thinking along

these lines of languages having an impact on each other has reemerged

under new labels such as language transfer, cross-linguistic influence
and substrate influence. Odlin (1989) defines transfer as follows:
‘Transfer is the influence resulting from the similarities and differences
between the target language and any other language that has been pre-
viously (and perhaps imperfectly) acquired’ (p. 27). He further distin-
guishes between borrowing transfer (the impact of the first language
on the second) and substrate transfer (the impact of the second language
on the first).
One of the most influential articles on cross-linguistic influence (CLI) is
probably Kellerman’s 1995 article ‘Crosslinguistic influence: transfer to
nowhere?’ He states:
At its simplest, the L1 can be seen as a direct cause of erroneous perfor-
mance, especially where such performance is shown to vary system-
atically among learners with different L1 backgrounds.
(Kellerman 1995, p. 125)
For Kellerman, deviation from some norm appears to be a crucial aspect of

CLI. Later on he adds:
In the give and take of unplanned conversation, where there may not be
sufficient attentional resources to devote to perfect monitoring of
linguistic performance or to the ongoing generation of long and detailed
circumlocutions … the learner can have recourse to ‘quick and dirty’
alternatives, among which are those based on L1 lexis.
(Kellerman 1995, p. 129)
The notion of insufficient resources for perfect monitoring assumes the

existence of fixed knowledge somewhere in the system that can be used to
check language production. This reveals a kind of thinking that is very
much based on a strict competence/performance distinction: there is
underlying knowledge and this may not be applied correctly. In the per-
spectives taken by Hopper (1998) and Larsen-Freeman and Cameron (2008)
such a distinction is no longer seen as valid. In the definitions and the
explanations given, it is unclear to what extent the most recent definition
of multi-competence still holds on to the competence/performance dis-
tinction as used in Chomskyan linguistics.
To summarize this part: in the tradition starting with Weinreich’s 1953
Languages in Contact up to Odlin’s 2009 article on transfer, the thinking
behind it is based on the ‘monolingual’ assumption:
– there are separate language systems that have an impact on each other,
– languages exist as stable entities in our brain.
The real question from an emergentist perspective is whether belonging to

a certain language is a defining and processing-relevant characteristic of a
linguistic element.

6.2.2 Languages as entities: psycholinguistic perspectives

What evidence is there for stable and language-specific processing and
storage?
There is a wealth of literature, mainly on the bilingual lexicon, in
which various experimental techniques have been used to study bilin-
gual processing. The discussion has been defined in terms of selective
versus non-selective access (Laheij 2005). The consensus now is that
the lexicon is organized in a non-selective way according to subset
hypothesis of Paradis (2004, 2009): in the lexicon, subsets are formed
based on use. Since words of a given language tend to be used
together they form a network. For speakers who often code-switch,
subsets may develop that consist of words that come from different
languages according to an external norm. It should be noticed that for
that particular speaker words need not be defined as belonging to one
language or another. The model of Paradis’ model is clearly usage-
based and emergent. What the elements in the lexicon are is not
entirely clear: it may be single words, but more likely there will be
larger units that are ‘sedimented’ on the basis of frequency of use. It
may well be that ‘words’ don’t have a separate status in the language
system at all, since they are hardly ever used in isolation, which
suggests that the concept of word is an artifact of learning a written
script with word spaces (Aronoff 1992).
The idea of subsets is not limited to the lexicon: the same principles may
be at work for syntactic or phonetic patterns. As argued in de Bot (2004),
there may be links between elements at different levels that may coacti-
vate each other: a sound that is associated with a specific language may
activate elements that ‘belong’ to that language.
One of the continuing discussions is about whether elements are
labelled for language. Laheij (2005) argues that if the conceptual specifica-
tion is detailed enough, no language labels are needed, while Hartsuiker
and Pickering (2007) present a model in which the language tag is an
integral part of the conceptual specification. From an emergentist perspec-
tive, it is unclear why there should be explicit language tagging, since
co-occurrence and associations of linguistic elements with specific set-
tings and interlocutors would suffice to lead to the selection of the right
words and thus the right language. This is the general mechanism: lan-
guage elements are encountered in specific settings and stored as such
along with episodic information on how and when similar settings will
lead to the activation of related elements. Elements will thus be associated
with communicative settings. Elements can be labelled consciously as
belonging to a specific language set but language as such does not act as
a cue in selection. For monolinguals the associated linguistic elements
will come from one language, for a regular code-switcher from two or
more languages. But again, such labelling is only done post hoc and at a
metalinguistic level and it is not necessarily a selection criterion.

The next question then is: does code-switching exist? If there is only one
merged system, and the speaker uses those elements that are associated
with the setting, there is no switching in the proper sense, since the tasks
are no different.
The fact that elements from two languages are used does not mean that
there are separate languages in the system of a code-switcher: she simply
takes the elements that are most appropriate and accessible. To quote
Hopper again:
Language is not a general abstract possession that is uniform across the
community, but is an emergent fact having its source in each individual’s
experience and life history and in the struggle to accomplish successful
communication. (Hopper 1998, p. 151)
6.2.3 The neurolinguistic evidence for separate languages

in our brain
A simple solution to solve this issue would be to find out whether separate
languages have their own neural substrate in the brain. There are two
sources of information about this: one is the evidence from aphasia, the
other is data on neuro-imaging.
For the latter, early studies (such as Kim et al. 1997) suggested that there
are different substrates for L1 and L2, but more recent studies have shown
that the findings were in fact caused by other factors. Stowe concludes her
review of this literature by stating: ‘There is no consistent qualitative
difference between the neural architecture supporting processing of
(the) two languages’ (2006, p. 305). In her more recent overview Sabourin
(2014) arrives at the same conclusion.
Various factors have been assumed to lead to different localizations,
in particular age of acquisition and level of proficiency, but the argu-
mentation has changed: it is likely that anything acquired early is
represented differently compared to what is acquired late, and that
applies to language as it does for other aspects of memory and cogni-
tion. Differences in proficiency reflect frequency differences that are
likely to have an effect on representations. So what leads to differ-
ences in processing and localization may be the effect of age of
acquisition and frequency and tells us little about differences in
neural substrates for different languages. As Paradis has been arguing
for quite some time, differences in proficiency may lead to other
strategies being used, which may be reflected in different brain
areas showing activation, but again, that is not about the localization
of L1 or L2 in itself.
Views on representations and their relation to brain structures have
changed considerably over the last decade. Hagoort (2006) summarizes
the earlier views as follows:

Architectural differences in the brain structure are indicative of func-

tional differences and, conversely, that functional differences demand
differences in architecture. (Hagoort 2006, p. 93)
In other words: differences in processing reflect differences in representa-

tion. Even stronger: because there are differences in processing, there
must be differences in representations. Following Hagoort again, the
new views on cognition and the brain focus on the plasticity of the brain
on the basis of input:
Functional differences between brain areas are in this perspective mainly

due to variability of the input signals in forming functional specializa-
tions. Domain specificity of a particular piece of cortex might thus not so
much be determined by heterogeneity of brain tissue, but by the way in
which its functional characteristics are shaped by the input.
(Hagoort 2006, p. 94)
This means that use shapes the brain and modularity is not innate but
emergent: due to repeated and associated use, certain brain areas will
show module-like behaviour. In the same way it could be argued that
different languages in the brain are emergent. Through associated use
networks will emerge that represent a given language, but these networks
are constantly changing and highly individual, because individuals’
experiences and contacts with the language will vary.
So will we ever find the localizations of our languages? It follows from
the argumentation given earlier that this is very unlikely. There is no
stable substrate, only unstable and constantly varying networks without
language labels. Different settings will activate different language forms
that have different substrates. It is therefore pointless to continue to try
to find the location of different languages. There is an interesting parallel
with an earlier discussion in the area of the bilingual brain. For quite
some time it was assumed that in bilinguals the right hemisphere plays a
more prominent role in language processing than in monolinguals.
Paradis (1990), in a biting commentary, has compared this line of
research with the search for the Loch Ness Monster: it must be there,
we just have to search harder even in the most unlikely places. But we
will never find it. Looking for languages in the brain may be the new Loch
Ness Monster.
As mentioned earlier, data from bilingual aphasia may provide us
with additional information about the processing of multiple
languages. There is a whole set of studies that might be interpreted
as providing counter-evidence for the assumption that there are no
separate and parallel language systems in the bilingual brain. These
studies report on cases of patients that show selective decline of
recovery (see Albert and Obler 1978 and Paradis 1987 for historical
overviews). Several patterns of selective decline and/or recovery have
been reported:

– loss of Lx not Ly,

– recovery of Lx but not Ly,
– cases of uncontrolled switching.
In these studies no damage to specific neural substrates has been found

consistently, so there doesn’t seem to be one particular brain area that
controls these processes. One explanation that has been offered is that in
many of those cases the patients also showed a pragmatic deficit (Ansaldo
and Marcotte 2007). Paradis (2009) interprets these data in a larger frame-
work of declarative and procedural knowledge in language use. Pragmatic
deficits of different kinds, that may lead to inadequate selection of lan-
guages, reflect a problem in procedural memory.
6.2.4 Evidence from second language development

As mentioned earlier, the idea of separate languages in the mind of the
second language learner has been generally accepted in the field of applied
linguistics. The use of terms like transfer and interference reflects this kind
of thinking. Data from very early writers in L2 (Dutch learners of English in
the first grade of secondary education; Verspoor, Schmid and Xu 2012) show
that, at least for this very early stage, it is difficult to talk about two language
systems. Below are short stories by two early foreign language learners:
Hello dis is my school he staat in apeldoorn. he is very big ai have very veel
teachers op my school and ik have er to very veeler zijn to very veel kids op
dese school de teachers and de kids walking door de school en have very
veel lol de englisch teachers says enlisch
(Hello, this is my school, it is located in the city of Apeldoorn, it is very big

and I have very many teachers at my school, and there are very many kids
at this school the teachers and the kids walk through the school and have a
lot of fun the English teachers speak English.)
Hello, i am Arnoud. I sit op the grammar school the Driestar College. I

found it well funny. I have nu veel more homework dan first. I hate english
and techniek. It are very crazy teachers. We have many leerlingen in the
new klas. I have veel friends. I moet heel veel biken to school. It is
ongeveer ten tot vijftien km biken.
(Hello, I am Arnoud, I am at the Grammar school named het Driestar

College. I found/find it nice. I now have a lot more homework than before.
I hate English and Technique. The teachers are very crazy. We have many
pupils in the new class, I have many friends. I have to ride by bike a lot to
get to school, it is about 10–15 km to my school.)
Data such as these cast at least some doubt on the idea that there are
separate systems in early learners. There can be no doubt that these young
writers intend to write in English, but they will be well aware of the fact
that what they actually write is a mix. But it is the best they can do.

6.2.5 A historical look at languages as entities

The history of language and its study is far beyond the scope of the present
chapter. Here some relevant notions will be touched upon only. It has been
argued that language labels were invented in the nineteenth century to
support the idea of the nation state: one state, one language (Pennycook
2004; Shohamy 2006). It could also be argued that dialectology developed
as a result of such thinking to curtail the embarrassing chaos of ways of
speaking, ordering different variants under larger headings. What was
needed was a standard that defined a language:
Standardization means that languages need to be used in certain ways, often
in sharp contrast to how they are actually used by people, especially with
regard to oral varieties, which are known to vary greatly from one person to
another and from one speech community to another. (Shohamy 2006, p. 64)
Makoni and Pennycook (2007) argue that languages can be seen as a

western invention to define ethnolinguistic groups, while the concept of
distinct languages may be far from them. Mansour (1993) notes that along
similar lines, language labels have been used to split groups that were
basically the same. The use of language labels made languages into entities
that could be defined and categorized. This fitted in with the craving of
linguistics to be taken seriously as a science. By defining languages, lin-
guists could enter the realm of science in which entities can be studied
objectively and organized and categorized as natural phenomena, just like
trees or insects.
6.3 Final remarks
In this contribution it is argued that we need to reconsider our thinking about

languages. In fact, ‘languages’ are cultural artefacts and there may be no
‘languages’ in our brain at all, only one merged system of situation-specific
utterances. Arguments for this position have been taken from linguistics,
aphasia studies, neuro-imaging studies, and studies on code-switching and
second language development. What is presented here is highly speculative
and needs more thinking and empirical support. To a certain extent it
challenges some of the assumptions of multi-competence, such as the exis-
tence of separate languages and the competence/performance distinction
that seems to persist in a muted form in even the most recent definitions.
Notes
1. The term Dynamic Systems Theory will be used here to refer to a group
of theories including complexity theory and theories about complex
adaptive systems that share a number of core characteristics, such as

dependence on initial conditions, iterative development, attractor

states and interconnectedness.
2. ‘I believe theories of bilingualism that entail the existence of parallel
but separate codes are misconceived’ (Hopper, personal communica-
tion, 24 February 2009).
References
Albert, M. and Obler, L. 1978. The Bilingual Brain. New York: Academic Press.
Ansaldo, A. and Marcotte, K. 2007. ‘Language switching in the context of
Spanish–English bilingual aphasia’, in J. G. Centeno, R. T. Anderson and
L. K. Obler (eds), Communication Disorders in Spanish Speakers, pp. 12–21.
Aronoff, M. 1992. ‘Segmentalism in linguistics: the alphabetic basis of
phonological theory’, in P. Downing, S. D. Lima and M. Noonan (eds),
The Linguistics of Literacy, pp. 71–82. Amsterdam: Benjamins.
Behrens, H. 2009. ‘First language acquisition from a usage-based perspec-
tive’, in K. de Bot and R. Schrauf (eds), Language Development over the
Lifespan, pp. 19–39. New York: Routledge.
Chan, H., Lowie, W. and de Bot, K. 2015. ‘Input outside the classroom and
vocabulary development: a dynamic perspective’, in M. Reif and
J. Robinson (eds), Culture and Cognition in Bilingualism. Philadelphia: John
Benjamins.
De Bot, K. 2004. ‘The multilingual lexicon: modelling selection and con-
trol’, International Journal of Multilingualism 1, 17–32.
De Bot, K., Gommans, P. and Rossing, C. 1997. ‘First language loss in a
second language environment: Dutch immigrants in France’, in S. Kroon
(ed.), Dutch Overseas: Studies in Maintenance and Loss of Dutch as an Immigrant
Language, pp. 153–162. Tilburg University Press.
De Bot, K., Lowie, W. and Verspoor, M. 2007. ‘A dynamic view as a comple-
mentary perspective’, Bilingualism, Language and Cognition 10 (1), 51–55.
Erard, M. 2012. Babel No More: The Search for the World’s Most Extraordinary
Language Learners. New York: Free Press.
Goles, E., Morvan, M. and Duong Phan, H. 2002. ‘Sand piles and order
structure of integer partitions’, Discrete Applied Mathematics 117, 51–54.
Hagoort, P. 2006. ‘What we cannot learn from neuroanatomy about lan-
guage learning and language processing: a commentary on Uylings’, in
P. Indefrey (ed.), The Cognitive Neuroscience of Second Language Acquisition,
pp. 91–97. Malden: Blackwell.

Hartsuiker, R. and Pickering, M. 2007. ‘Language integration in bilingual

sentence production’, Acta Psychologica 128 (3), 479–489.
Hopper, P. 1998. ‘Emergent grammar’, in M. Tomasello (ed.), The New
Psychology of Language: Cognitive and Functional Approaches to Language
Structure, vol. 1, pp. 155–176. Mahwah, NJ: Lawrence Erlbaum.
Kellerman, E. 1995. ‘Crosslinguistic influence: transfer to nowhere?’,
Annual Review of Applied Linguistics 15, 125–150.
Kim, K., Relkin, N., Lee, K. and Hirsch, J. 1997. ‘Distinct cortical
areas associated with native and second language’, Nature 388 (9938),
171–174.
Laheij, W. 2005. ‘Selection processes in monolingual and bilingual lexical
access’, in J. F. Kroll and A. M. B. de Groot (eds), Handbook of Bilingualism:
Psycholinguistic Approaches, pp. 289–304. Oxford University Press.
Larsen-Freeman, D. and Cameron, L. 2008. Complex Systems and Applied
Levelt, W. J. M. 1989. Speaking: From Intention to Articulation. Cambridge, MA:
MIT Press.
Makoni, S. and Pennycook, A. 2007. Disinventing and Reconstituting Languages.
Mansour, G. 1993. Multilingualism and Nation Building. Clevedon:
Odlin, T. 1989. Language Transfer: Cross-Linguistic Influence in Language
Learning. Cambridge University Press.
Odlin, T. 2009. ‘Methods and inferences in the study of substrate influ-
ence’, in M. Filppula, J. Klemola and H. Paulasto (eds), Vernacular
Universals and Language Contacts, pp. 256–279. New York: Routledge.
Paradis, M. 1987. The Assessment of Bilingual Aphasia. Hillsdale: Erlbaum.
Paradis, M. 1990. ‘Language lateralization in bilinguals: enough already!’,
Brain and Language 39, 576–586.
Paradis, M. 2004. A Neurolinguistic Theory of Bilingualism. Amsterdam and
Philadelphia: John Benjamins.
Paradis, M. 2009. Declarative and Procedural Determinants of Second Languages.
Amsterdam and Philadelphia: John Benjamins.
Pennycook, A. 2004. Language as a Local Practice. New York: Routledge.
Sabourin, L. 2014. ‘fMRI research on the bilingual brain’, Annual Review of
Shohamy, E. 2006. Language Policy: Hidden Agendas and New Approaches. New
York: Routledge.
Smith, K., Kirby, S. and Brighton, H. 2003. ‘Iterated learning: a framework
for the emergence of language’, Artificial Life 9, 371–386.
Spivey, M. 2007. The Continuity of Mind. New York: Oxford University Press.
Stowe, L. 2006. ‘When does the neurological basis of first and second
language processing differ? Commentary on Indefrey’, in P. Indefrey

(ed.), The Cognitive Neuroscience of Second Language Acquisition, pp. 305–311.

Malden: Blackwell.
Van Geert, P. 1994. ‘Vygotskian dynamics of development’, Human
Van Geert, P. 2008. ‘The dynamic systems approach in the study of L1 and
L2 acquisition: an introduction’, Modern Language Journal 92, 179–199.
Van Geert, P. 2011. ‘The contribution of complex dynamic systems to
development’, Child Development Perspectives 5 (4), 273–278.
Verspoor, M., Lowie, W. and de Bot, K. 2010. A Dynamic Approach to Second
Language Development: Methods and Techniques. Amsterdam/Philadelphia:
John Benjamins.
Verspoor, M., Schmid, M. and Xu, X. 2012. ‘A dynamic usage based per-
spective on L2 writing development’, Journal of Second Language Writing 21
(3), 239–263.
Weinreich, U. 1953. Languages in Contact: Findings and Problems. New York:
Linguistic Circle of New York Publications.

7
Dominant Language
Constellation
Larissa Aronin
7.1 Introduction
The perspective of multi-competence became extremely useful when the

societal awareness of bilingualism generated a demand for understanding
what it means to be a bilingual. This was a welcome response to the chal-
lenges encountered by non-native speakers, mostly of English, by language
teachers, scholars and all those associated with second language acquisition
and use. Initially defined as ‘knowledge of two or more languages in the
same mind’ (Cook 1991, p. 103; Cook website) and ‘the compound state of
mind with two grammars’ (Cook 1992, pp. 557–558), it now ‘involves the
whole mind of the speaker, not simply their first language, or their second’
(Cook 2013a, p. 3768) and ‘the overall system of a mind or a community that
uses more than one language’ (Chapter 1, this volume).
The multi-competence standpoint is widely accepted these days. It has
become especially relevant with the development of multilingualism
research. Bilingualism and multilingualism studies are increasingly
departing from linear and simplified explanations, instead adopting
holistic, dynamic and complexity approaches. And this change of perspec-
tive is not an arbitrary mind-game for researchers. Rather, it reflects the
crucial changes that have taken place in the world: technological break-
throughs, social reordering, accompanied by changes in ideologies,
perceptions and the inevitable transformation of identities and social
practices of language use and acquisition.
The distinctive character of the current global sociolinguistic situation,
which has become an inherent and salient property of contemporary
society, was emphasized in the concept of New Linguistic Dispensation
(Aronin, Fishman, Singleton and Ó Laoire 2013; Aronin and Singleton
2008, 2012).

Multi-competence and Dominant Language Constellation 143
The multi-competence perspective is valuable in that it enables one to

consider the abilities of bi- and multilinguals and their complex and
mutable identities in the super-diverse global environment. As the inves-
tigation of multilingualism in education and social contexts unfolds, the
concept of multi-competence proves increasingly useful in facilitating
research and further understanding of the growing number of multilin-
guals and multilingual communities.
Providing insight into the essential difference between monolingual
native speakers and L2 users (including learners), the perspective of
multi-competence is, at the same time, easy to understand and, what is
not to be disregarded, attractive to non-native speakers, who comprise a
majority in the contemporary world. The concept of multi-competence
reveals the nature of bi- and multilinguals as essentially different from
only-one-language-speakers, in that ‘it assumes that someone who knows
two or more languages is a different person from a monolingual, and so
needs to be looked at in their own right rather than as a deficient mono-
lingual’ (Cook 2013a, p. 3768).
In this chapter I will invite the reader to consider the concepts and
perspectives which signify the relatively recent shift in traditional norms
of viewing language speakers, language acquisition and use. The concept
of Dominant Language Constellations (DLCs) will be described in some
detail, with the aim of demonstrating how multi-competence is expressed
and can be studied via DLCs.
7.2 Multi-competence and related concepts
Before approaching the inter-relation between the perspective of multi-

competence and the concept of DLC, the focus of this chapter, I will briefly
describe the concepts of multilinguality and language repertoire and show how
they relate to each other and to the two concepts under discussion.
Language repertoire. The concept evolved from a monolingual
perspective, and under the name of ‘verbal repertoire’ it was deployed by
Gumperz (1964, pp. 137–138), who referred to the sum of various skills in
one language. Later it was expanded to include more than one language, as
‘an individual’s particular set of skills (or levels of proficiency) that permit
him or her to function within various registers of (a) language(s)’ and the
‘totality of linguistic varieties shared by the group as a whole’ (Schiffmann
1996, p. 42). Pütz expressed this extension of the notion of linguistic
repertoire from monolingual situations to bilingual ones in the following
words:
‘Repertoire’ bezeichnet nicht nur die Kompetenzen monolingualer
Sprecher, sondern auch im bilingualen Kontext sind die Aspekte des
Sprachwechsels (Code-Switching) wie auch der Sprachwahl (language choice)

144 LARISSA ARONIN
bedeutsam . . . Der Begriff ‘Repertoire’ lässt sich sowohl auf den indivi-
duellen Sprecher wie auch auf die sprachlichen Fertigkeiten einer
Sprechgemeinsamkeit anwenden. (Pütz 2004, p. 227)
(‘Repertoire’ not only designates the competencies of monolingual speak-

ers, but in addition, in bilingual contexts the aspects of codeswitching and
language choice come into play . . . The concept of ‘repertoire’ is applicable
both to the individual speaker and to the language capacities of a collec-
tivity sharing a language commonality.) (translation by David Singleton 2012)
For the argument in this chapter, it is important to point out that in the
notion of language repertoire, whether taking into consideration entire
languages or zooming in on various skills in each language, the stress is
on the sum of available language varieties and/or skills. Repertoire
includes every language or skill that an individual or a group possesses.
As opposed to that, later in this chapter we will see that Dominant
Language Constellation differs from the language repertoire, in that the
DLC includes not all, but only those languages or skills selected for their
prime importance.
Multilinguality. Another concept related to the discussion of multi-
competence and DLC, is multilinguality (Aronin and Ó Laoire 2001, 2004).
Multilinguality is an epitome of multi-competence. Each multilingual has
his/her own multilinguality, that is, a particular identity defined by his/her
language or languages to a great extent. As a form of identity, multilin-
guality is different from bilinguality (Hamers and Blanc 1983) and from the
kind of identity typical of monolinguals. The linguistic component plays a
crucial role in shaping identity. And this is understandable, since language
is among the most essential attributes of an individual. Human language
provides informational, cognitive and emotional input, and through this
forms and transforms a person. One’s identity is also expressed mostly
through language.
The original definition of multilinguality was:
an individual store of languages at any level of proficiency, including
partial competence and incomplete fluency, as well as metalinguistic
awareness, learning strategies and opinions, preferences and passive or
active knowledge of languages, its use and learning/acquisition.
(Aronin and Ó Laoire 2004, pp. 17–18)
In the first place, multilinguality comes out of linguistic behaviour; it

includes idiosyncrasies and peculiarities of communicators, as well as
individual disabilities and blessings, such as physical abilities of articula-
tion, being a sign language user, dyslexic or having a perfect pitch for
singing.
The phenomenon of multilinguality is far from being a purely linguistic
characteristic. It incorporates various other human aspects or dimensions,
from the physiological and psychological, to psychological and social.

Multilinguality is also about cognitive behaviour and acquisition processes,

features of age, gender and memory. Beyond the knowledge of two or
more languages and cognitive specifications, the concept of multilinguality
necessarily includes everything that causes and accompanies this knowledge. It
embraces everything that results from using and learning several languages,
both in the present and also potentially in the future. Thus, multilinguality
is an individual characteristic, based on one’s linguistic assets, abilities and
experience, and expressed through actions, perceptions, attitudes and
personal life scenarios, both real and possible.
From such a perspective, multilinguality is a psychological notion, in
that it describes the characteristics of an individual. It is also sociolinguis-
tic, as it traces the social status, opportunities and involvement of a multi-
lingual. Investigating multilinguality, one has to regard the social milieu
as an important influence, and consider the societal consequences of
language learning and use. Each person’s multilinguality is shaped by
legacies of historical events and family history, embedded assumptions
and social opportunities. Related to social, family and career activities,
educational awareness and lifestyle, as well as the range of accompanying
and pervading emotions, affective states (directly or tangentially con-
nected with or through language), one’s multilinguality stems from, and
influences, the social and private life of an individual.
There is a remarkable interplay between the multiple, closer or tangen-
tially related identity facets, such as nationality, gender, age and the self as
a citizen, but multilinguality is seen as a whole, not divided or separated
into distinct sub-identities. These identity facets are coloured, modified,
stimulated, excited, forced to the forefront, or hidden away by language-
related factors. Each individual’s multilinguality is unique.
Here is how Humphrey Tonkin, whose language repertoire includes
Esperanto, Latin, English, French, German, Italian and Dutch, describes
the influence that one of his languages exerted on his identity:
my circle of Esperanto friends and acquaintances, and indeed the
Esperanto community in general, have a broad if incalculable effect on
my knowledge and views of the world . . . Esperanto has shaped my world [my
italics – L. A.], expanded my linguistic skills, left me richer in experience,
and, I would hope, made me politically more aware that would ever have
happened otherwise. (Tonkin 2009, p. 206)
How is multilinguality related to multi-competence? Multi-competence

is a largely linguistic perspective, which emphasizes the specific qualities
and abilities of the whole group of individuals – those who speak more
than one language. It celebrates the ability and the quality that all bi- and
multilinguals share. Multilinguality is a unique personal characteristic for
tri- and multilinguals, the foundation of which is multi-competence. It is
multi-competence that shapes the identity of a multilingual in accordance
with each particular environment.

146 LARISSA ARONIN
Table 7.1 Inter-related perspectives and concepts
Concept or perspective What it is about Emphasis applies to
Multi-competence Ability and as such charac- whole mind/cognition

teristics of the individual
Multilinguality Type of identity expressed/ whole identity
enacted via languages
DLC Set of selected languages selected languages
and skills
Language repertoire Totality of languages and totality of languages or skills
skills of an individual or
community
The points of similarity and difference are arranged in Table 7.1.

All the concepts in this table refer to the use of two and more languages.
They all describe and characterize the situation of bi- and multilingualism,
and refer to bi-multilingual situations as distinct from monolingual. Most
of them emphasize (or imply, as in ‘language repertoire’) the emerging
quality that appears from the use of two and more languages. The concepts
of multi-competence, multilinguality, DLC and language repertoire (Table 7.1)
nevertheless differ in focus, viewpoint and scope.
Some of these notions refer equally to a society and an individual, but all
of them are very important to the individual and identity. Some are equally
relevant to both, such as language repertoire, and others are more relevant
to the individual’s multi-competence, or only to the individual’s multilin-
guality. DLC is the concept that can be applied both to an individual and to a
community.
Multi-competence is a fundamental perspective enabling us to under-
stand the unique abilities of more-than-one-language speakers. But bilin-
gual and multilingual speakers and communities are many. How can one
understand the multi-competence of a community or an individual more
exactly, when relating to each particular case? Here is where the DLC
concept comes into play. In the following I will first describe the concept
of DLC and then show how multi-competence and DLC are connected.
7.3 The concept of Dominant Language Constellation (DLC)
A Dominant Language Constellation, that is, the constellation of one’s

dominant languages, is a group of one’s most important (vehicle) lan-
guages, functioning as an entire unit, and enabling an individual to meet
all his or her needs in a multilingual environment (Aronin 2006; Aronin
and Ó Laoire 2004; Aronin and Singleton 2012). The Dominant Language
Constellation includes only the most expedient languages for a person,
rather than all the languages known to them, as would be the case in

Dutch Finnish Swedish Portuguese Swahili Putonghua Africaans
Amharic
Formal
Yiddish
Arabic Latin
Polish
Brasilian
DLC Portuguese
Spoken Arabic
Italian English
Hebrew Spanish
Rose
Figure 7.1 DLC in Israel: Rose, Arabic sector school (Aronin and Jessner 2014)
Key. 5-point stars: languages of the language repertoire. Orbiting 5-point stars: repertoire
languages with weaker knowledge or seldom used. 4-point stars: languages a person is
exposed to in their close environment and often understands – whether separate words and
phrases, or more. It can be either a heritage language in a family or a language often heard
due to social proximity, whether by wish or by circumstance.
language repertoire. Unlike a language repertoire, a DLC comprises the

languages which, together, perform the most vital functions of language.
Consider for example, an account of the languages of Rose, an English
teacher in an Arab sector school in Israel (see Figure 7.1). Rose reported
mastery of eight languages. Due to diglossia, she discriminates between
Formal Arabic and Spoken Arabic, and thus her language repertoire com-
prises Spoken Arabic, Formal Arabic, Hebrew, English, Spanish, Brazilian
Portuguese, Italian and Latin. In Formal Arabic, mostly used at school,
Rose’s proficiency is excellent: she can use it to write formal letters and
carry out official school matters. In addition, Rose can speak and compre-
hend Latin to a limited degree; being a Catholic, she listens to Mass in
church on Sundays and religious holidays. Italian, Spanish and Brazilian
Portuguese, which she learned on her own, serve Rose mostly for enjoy-
ment (songs, TV programmes) and travel. Rose mastered comprehension
and speaking skills in Spanish quite well, and also manages some basic
writing. She can understand some Italian and she also has sufficient
receptive skills in Brazilian Portuguese.
This list of skills in a number of languages is Rose’s language reper-
toire. Her Dominant Language Constellation consists of only three
languages out of the eight: Spoken Arabic, English and Hebrew (see

148 LARISSA ARONIN
Amharic
Formal
Yiddish
Arabic Latin
Polish
DLC Spanish
Spoken Arabic
French Hebrew
English
Iman
Figure 7.2 DLC in Israel: Iman, Arabic sector school (Aronin and Jessner 2014)
Figure 7.1). Hebrew and Arabic are the official languages of Israel.
Spoken Arabic is her mother tongue, and the language Rose has mas-
tered best to use in any situation. It is used to communicate within the
family and in the close community. Rose uses Hebrew in order to
communicate with people in Israel who do not know Arabic. English
and Hebrew are also the languages of her studies at college. Rose uses
English at work, reads books in English, searches the internet, for
texting and sometimes with her children. In her daily life Rose mostly
relies on these three languages.
Rose’s colleague, Iman, who is a member of the same community and is
also an English teacher, selected the same three languages for her DLC:
Spoken Arabic, Hebrew and English. The languages and skills of Iman’s
repertoire are similar to Rose’s, that is, Formal Arabic as a separate lan-
guage variety from Spoken Arabic, used only in specific situations and, like
Rose, Iman has mastered some Latin for religious purposes and Spanish for
enjoyment. In addition, Iman listed listening, speaking, reading and writ-
ing in French (Figure 7.2).
The following description is of Liti, an Israeli English language teacher at
a big school with a Jewish population. Her DLC diverges from the previous
two, in that it involves Hebrew, English, Romanian and German
(Figure 7.3). As languages in her repertoire, Liti listed Yiddish with good
comprehension skills, Hungarian, a language spoken by Liti’s husband and
his parents, and French, which she used to study earlier in her life. Thus
her repertoire includes seven languages.

Dutch Italian Spanish Portuguese Swahili Putonghua Africaans
French Arabic
Yiddish
Russian
DLC
Hebrew Amharic
English
Romanian
German
Hungarian
Liti
Figure 7.3 DLC in Israel: Liti, Jewish sector school (Aronin and Jessner 2014).
For key see Figure 7.1
Out of the seven languages of her repertoire, Liti uses four vehicle
languages to satisfy most of her needs that human language is respon-
sible for. She uses Hebrew with her children, students, colleagues and
for all formal events and needs. She uses English for her academic
studies and at school, as well as for international exchange. Thus, two
languages of Liti’s DLC are the same as those of her Arab colleagues,
and she uses these languages in the same situations as Rose and Iman.
But the rest of her DLC diverges. The other two languages of the
constellation reflect Liti’s origin, her own life trajectory and her parti-
cular needs. Liti grew up as a bilingual in a Romanian–German-speak-
ing community; these two languages were spoken by a Jewish minority
in the north of Romania, in the Bucovina region. Now these two
languages ensure the important personal aspects of Liti’s life, and
preserve emotional and historical ties. She uses Romanian to speak to
her husband and her sister, and her proficiency both in writing and in
oral skills is excellent. Finally, she uses German, her heritage language,
to communicate with her mother (Figure 7.3). These are not the only
possible DLCs in multilingual Israel. For example, a member of a
Russian-speaking community may speak Russian, Hebrew and English
(Figure 7.4).
Radmila Popovic’s (2009) narrative about her own languages encapsu-
lates the picture of language roles, especially as she also uses the ‘space’
metaphor as in the figures above.

150 LARISSA ARONIN
Amharic
Latin Yiddish
Spanish
Polish
Arabic
DLC
Hebrew
French
Ukrainian Russian
English
German
Figure 7.4 DLC in Israel: user of Russian, Hebrew and English (Aronin and Jessner 2014)
My languages peacefully coexist and actively interact in my mind. They

feel to me like a constellation of permeable celestial bodies, which come
closer or move away from each other travelling along their own orbits.
In this tiny galaxy the central position is taken by Serbian and English.
Most of the time these languages are interconnected, then occasionally
merge, when I switch from one to another, and also grow apart – when for
a while I happen to use one language more frequently than the other.
Russian orbits a distance away from them, but maintains direct airlines
with the central bodies. French and Greek are barely visible, dwarf planets
with little solid matter, subsisting at the very edge of the system.
(Popovic 2009, p. 46)
Most countries today include multilingual populations. For example,

among the individual DLCs in Australia we can find English/Polish/
German, Latvian/English, and Bosnian/German/ English. In Argentina
there is a population whose DLC consists of Welsh/English/Spanish; and
in Hungary, someone’s vehicle languages can be Romanian/Hungarian/
German.
The concept of Dominant Language Constellation is relevant not only
for individuals, but also for communities. Groups characterized by a com-
mon DLC may share territorial, social, ethnic and/or business character-
istics. In Malta, regular use of English and Maltese is often associated with
families pertaining to a higher socioeconomic class (Caruana and
Lasagabaster 2013); these two languages characterize this particular seg-
ment of the population. The educational system supports the use of

English and Maltese, and the most popular choice in additional languages
at school is Italian; multiple TV channels provide the input of Italian in
addition to the historical tradition of its use in this country (Caruana and
Lasagabaster 2013). One can assume that a DLC of English/Maltese/ Italian
is shared by a section of citizens in Malta.
An assortment of DLCs can coexist in the same physical space. It is
common under the new linguistic dispensation (Singleton, Fishman,
Aronin and Ó Laoire 2013) to have a number of DLCs in a country; in
Germany, for example, Polish/German, Polish/German/English, Turkish/
German, or Russian/German/English. In Switzerland, in the canton of
Ticino, there are people for whom English/German/French or Italian/
French are crucial. In a Romansh valley there is a small population of
quadrilinguals with Romansh/German/Italian/French. And these are only
some of the many DLCs deployed in Switzerland.
While the repertoire of multilinguals may include five, six, seven, eight
or more languages, the typical DLC consists of three languages. I am not by
any means suggesting that three is a magical or exact obligatory number; it
is simply a fact that out of many languages they know, people normally use
three (sometimes two, sometimes four). This number arises from empiri-
cal observations and studies on various aspects of multilingualism. A
relevant study, carried out in Ireland and Israel (Aronin and Ó Laoire
2005), demonstrated that, although multilingual participants reported
up to eight languages:
– the average number of languages that respondents perceived as neces-

sary for their social, cognitive and emotional needs was three;
– the most frequent number of languages reportedly spoken (as opposed to
languages known) according to these findings was three;
– even the languages in which the multilinguals in the study reported
communicating with their pets were languages from their DLC rather
than languages from their wider repertoire (Aronin and Ó Laoire 2005;
Aronin and Singleton 2012, p. 63).
Three languages in a DLC seems a plausible number, also taking into

account the following considerations: it is both impossible and excessive
to use all the languages and skills in all domains every day. The natural
tendency of humans to simplify and shorten an effort was formulated by
Zipf (1949), and is known as Zipf’s law, or the principle of least effort.
Therefore, in language practices, like any other kind of human behaviour,
in any given context, the number of languages tends to be reduced. On the
other hand, the objective necessity in several languages presses in the
opposing direction: the more languages that are in use, the more advan-
tages it gives to a person. The result of these forces acting simultaneously
but in conflicting directions is a certain number of languages, revolving
around three. This is how, without employing all the available language
skills, at each particular period in life, a user selects those skills which

152 LARISSA ARONIN
ensure the ability to carry out the functions of cardinal importance for
which human language is responsible.
Which languages (e.g. Swedish or Inuit) comprise a DLC depends primar-
ily on the social environment. The societal, political and historical circum-
stances of the country one lives in, and of a specific community, define the
composition of a DLC to a great extent. As the circumstances change, so do
DLCs. Millions of people living in Azerbaijan, Latvia and Belorussia used to
include Russian in their DLCs when it was ‘the language of international
communication’ in the USSR. In recent decades, Russian as a second or
foreign language has been replaced by English, German, Spanish or French
in the territories and polities of the former USSR. Constellation makeup
depends on human factors too: parents led by their life experiences,
ideologies and preferences arrange language acquisition for their children.
In times of mobility and openness, users themselves increasingly make
their own decisions, based on individual choice (on sociolinguistics as the
study of speakers’ choices see Coulmas 2013).
Rudolf Nureyev (1938–1993), the legendary Russian and British dancer
and choreographer, was born in the USSR to a Tatar mother and a Bashkir
father. From his early days, the future ballet star had at least three languages
and cultures to rely on: Tatar, Bashkir and Russian, a dominant language,
and later at the Leningrad Choreographic School, French. After breaking
through to the West, he added English as a language of communication in
his new environment. In reminiscences and memoirs of those who knew
him, Nureyev comes across as strong and independent. He made a decisive
selection of his own languages by allegedly just sticking to English in public.
The biographer Mario Bois recalled how Nureyev managed his
languages:
Он соглашался разговаривать только по-английски. Он знал немало других
языков, само собой разумеется, русский, неплохо говорил на французском и
итальянском, более или менее справлялся с немецким и испанским, знал кое-
что по-датски. Если вы заговаривали с ним по-французски, он отвечал вам на
английском. Он притворялся, будто знает один этот язык, что позволяло ему
‘подлавливать’ неосторожных. Не подавая виду, он слушал все, все схватывал.
(Bois 1993)
(He agreed to speak only in English. He knew a few other languages,

certainly Russian, was quite good in French and Italian, coped more or
less with German and Spanish and knew something from Dutch. If you
started speaking with him in French, he would answer you in English. He
pretended to know only this language, which allowed him to ‘catch’ the
careless. Without showing it, he would listen to everything, grasped every-
thing.) (My translation – L. A.)
The other side of the coin is indexicality, the first side of the coin being
dependence on a given social context, and some degree of personal choice,
as described above. Indexicality with reference to a DLC means that it
suggests the range of attributes of a person or a community, with some

degree of certainty. In the first place, the choice of languages indicates the
origins and ethnicity of their users, their geographical home; the combina-
tion of languages (e.g. Indo-European and Niger-Congo language families
in one DLC) discloses their life trajectory much more exactly, and in a
manner which is closer to life than a language repertoire – the particular
skills deployed, the proficiency level in each, the reasons and ways of
acquiring each of the languages. The prestige of the languages in a DLC
reveals a wealth of information about it as instrumental in finding a job
and in pursuing a career, as well as indicating one’s status in a particular
place, in a wider or closer community. This, and shifts that might take
place in a DLC (such as losing some skills, or improving others during an
individual’s lifespan), may help in defining a person’s profile.
The core advantage of the DLC concept is that the entire focus shifts
from the investigation of separate languages to the exploration of their
constellations. Researching language acquisition and use through a DLC
means ‘considering whole sets of languages as units, rather than focusing,
one by one, on the specific languages used by given individuals or groups’
(Aronin and Singleton 2012a, p. 69).
Cook (2013a, p. 3769) noted that ‘The commitment to the independent
L2 user in SLA research is often a matter of rhetoric rather than the reality
reflected in its research questions and techniques.’ The DLC approach
suggests taking into consideration all of the languages in the constellation
simultaneously, or, employing Ferdinand de Saussure’s concepts of syn-
chrony and diachrony (1916/1959) to study the use of several languages
synchronically. As mentioned elsewhere (Aronin 2006), in multilingual-
ism studies the concept of synchrony acquires additional significance, as it
may channel our attention to the differences, similarities, configuration of
skills and characteristics in each of the languages within a particular
constellation. According to Cook (Chapter 1, this volume), multi-compe-
tence concerns the total language system (L1, L2, Ln) in a single mind, and
the synchronic study of language constellations allows us to study multi-
competence in a way that departs from the sorry situation described
above.
The DLC approach shifts the study of language use and acquisition away
from the monolingual perspective, as it re-sets the point of departure from
one to several languages. DLC resonates with the idea of multi-competence,
as it evokes ‘visualizing’ the existence of more than one language in one’s
mind and activates interest in the complex interaction of languages.
A DLC, though consisting of a number of languages, can be construed as
one language unit. It is ‘a language’ in the meaning of ‘a human language’
and in the sense of ‘Language as a representation system known by human
beings’ (Cook 2013b, p. 27), as well as in the sense of ‘Language as a form of
action’ (Cook 2013b, p. 29).
The several languages of a DLC together carry out all language functions
that human language is responsible for: a means of communication and

154 LARISSA ARONIN
cognition, an expression of identity. In theory each language of a DLC can

alone perform all the possible functions for an individual or a community,
but this is rarely the case in the contemporary world. Global trends
(Aronin, Fishman, Singleton and Ó Laoire 2013; Aronin and Singleton
2008; Fishman 1998) dictate that, instead of one language, the language
faculty is exercised in different languages.
For an illustration of how several important languages can naturally
work as an entity, we can turn to a rich experience of India in multi-
lingualism, where such situations are common. Suresh Canagarajah
(2007, in MacKenzie 2012, pp. 84–85) notes that in multilingual commu-
nities in South Asia and Africa, simultaneous childhood acquisition makes
it hard to say which language comes first, or to identify a mother tongue or
native language. Canagarajah writes:
Such communities are so multilingual that in a specific speech situation
one might see the mixing of diverse languages, literacies and discourses. It
might be difficult to categorize the interaction as belonging to a single
language. (Canagarajah 2007, p. 931)
From the above we can conclude that a DLC has ‘two faces’, one as a unit
and the other as a complex system consisting of several components. The
‘unit’ feature is important for theoretical understanding and methodol-
ogy, while the fact that at the same time a DLC is a complex structure of
components provides a wealth of opportunities to research particular
multilingual individuals and communities.
Section 7.4 is devoted to elucidating the connection between the DLC
and multi-competence, and suggests some directions of investigation into
multi-competence using a DLC as a tool.
7.4 Multi-competence and DLC
7.4.1 DLC as an approach for research in multi-competence

Any particular DLC is an embodiment or reification of multi-competence
in real life. Or, putting it differently, multi-competence transpires through
activities with real languages. Every multilingual is endowed with multi-
competence but, as with other human characteristics, this one is mani-
fested in its own way with each individual via his/her own DLC.
In general, the DLC approach can enhance and focus theoretical under-
standing of multilingualism, specify and organize previously obtained
knowledge, and serve as a research methodology. The concept of the
Dominant Language Constellation can serve as an abstract model in
research into multi-competence and language acquisition (consider
Table 7.2).
The concept of DLC can be beneficial not only for theoretical under-
standing but also in practical investigation of particular DLCs. The

Table 7.2 The connection between multi-competence

and DLC and their role in research
Concept Multi-competence DLC
Concept type for research Perspective Model

Manifestation in real life Ability Languages
following points will elaborate on how a DLC approach can be useful for
research in multi-competence and multilingualism.
The particular languages of a dominant constellation make it possible
to outline and structure the ability which we call multi-competence.
While this set of dominant languages may change during one’s life,
accepting other languages and dismissing others, or moving some skills
to the foreground, according to the particular needs, the feature of multi-
competence still stays with a person, perhaps acquiring different shades
and hues, which, in turn, influences multilinguality. Thus, each particular
DLC is a representative form of one’s multi-competence.
In this vein, the DLC approach provides the framework within which
multi-competence develops. Therefore we may systematize the knowledge
accumulated on working with several languages together, in multilingual
situations in this framework. A DLC, perceived and investigated as a system,
which it certainly is, is manageable. In this way, it is possible to assemble
data, either in language policy, applied linguistics or language acquisition,
in a single resource, in a way that is efficacious for further analysis.
The DLC approach underlines the difference between multilinguals and
monolingual speakers, making it more transparent. This is clear, first of
all, through the qualities lacking in monolinguals, and their presence in bi-
and trilinguals. The two distinctive features of bilinguals that are denied to
monolinguals are translation and code-switching. As Cook noted, ‘The
monolingual native speaker is language-deprived; they would have
acquired multi-competence in more than one language if their caretakers
had not deprived them of a second language’ (Cook 2009, p. 156). Indeed,
viewing a DLC as a point of departure, we can register a lack of one more
important quality for monolinguals: they simply cannot have a DLC.
Furthermore, a DLC is more typical of tri- and multilinguals than of
bilinguals. For bilinguals the DLC will consist only of the two languages.
The distinctive nature of monolingualism, from both bilingualism and
tri- and multilingualism, becomes obvious from the following considera-
tions. If we have one entity for cognition, we can learn only about this
particular object or phenomenon, for example, acquiring French as a first
language would allow contemplating a single grammar, pronunciation
system and lexis. Knowing one language as a monolingual is connected
with many social and cognitive opportunities, but these are limited to this
language only.

156 LARISSA ARONIN
With two objects there is a new opportunity for cognition and compar-
ison, which is not available when we investigate one entity. On the basis of
two things we attempt to make predictions with some confidence. But with
three things, the possibilities of the human mind penetrating deeper into
the nature of things increase exponentially. Undoubtedly, three things
under study release a flood of ways to encounter, compare and categorize,
which is not possible with two.
DLC shows that measuring L2 users and L3 users against L1 users is
senseless since the nature of a ‘human language’ consisting of three
languages as a unit is characterized by constant and complex interplay
between its parts. The nature of the unit ‘human language’ which consists
of one language is less complex. A bilingual DLC would be different from a
trilingual and multilingual.
A DLC approach makes it easier to understand how societal history and
environment determine, or modify, the choice of languages, and thus the
ensuing quality or type of multi-competence. Thus multi-competence,
being largely a linguistic perspective, necessarily grows through the
human social tissue. This is because it is largely the societal order that
imposes languages on the individual, whether directly, through official
languages and language policies which promote and allow only certain
languages, and indirectly, via the overt or covert stimulation or discour-
agement via attitudes, and opportunities for jobs or status. The language
constellation that is dominant for a particular person is greatly dependent
on, and often imposed by, the constellation of languages in a community,
state or region of residence. Thus, the concept of DLC connects the indivi-
dual and the societal, as we will see later.
The key aspect of the multi-competence perspective is that the users of
an L2 (and L3, Ln) are language users in their own right (Cook 2013a,
p. 3768). DLC demonstrates and specifies how each multilingual user is
unique. Awareness of the uniqueness of each particular multilinguality
and DLC, which are both based on multi-competence, help one to appreci-
ate the super-diversity of language users in the world.
The next step would be to learn more specifically how multi-competence
‘works’ for various individuals and populations.
7.4.2 DLC as a method for research in multi-competence

Studying Dominant Language Constellations enables us to go into parti-
cular details and accumulate relevant real data about people and groups
endowed with multi-competence.
A personal DLC is an instance of the closest possible language contact,
literally in one person. The interaction of languages of the constellation is
what generates multi-competence and is a potent subject for research.
Studying each DLC as a unit consisting of a number of subsystems yields
a wealth of data. Multi-competence, a quality typical and unique for bi- and

multilinguals, has particular manifestations, which can be detected

through the lens of the DLC. In order to investigate the particularities of
particular DLCs, researchers could deal with its constitution, configuration
and dynamics (Table 7.3).
The constitution of a DLC is the data answering the questions ‘what?’
or ‘which?’ What languages comprise this DLC? Which skills are mostly
used? What is the learner’s age at the outset of studying L2 and L3? And
also, at what time of the day is a certain language used? In which domain?
In what language community?
A configuration of ‘whats’ in a DLC refers to the proportional data
regarding each language and its role in the entire unit of the DLC. A DLC is
a compound of distinct components, each of which may have a different
weight in relation to the other parts. Each language, being a component of
the unit (a DLC) has its own ‘weight’, or share in terms of the time being
used, the domain in which it is used, the level of proficiency, importance
for one’s career or soul. Competence may be very high in one language and
rather low in two others; the time of learning two languages can be from
birth, and the third language acquired later in life.
Finding frequent configurations may disclose patterns, typical configu-
rations of language use in a DLC, such as patterns of language mode usage
in a DLC. Some DLCs might be characterized by the predominant use of
one language, while others mostly deploy the parallel use of two or even
three languages in the case of the majority of the community being equally
multilingual. Another example of a typical configuration pattern might be
the correlation between high proficiency in the native language, less
comprehensive mastery of the L2, and partial competence in the L3 (see
more examples in the ‘Configuration’ column in Table 7.3).
The dynamics of a DLC refers to alterations in the constitution and
configuration of a DLC, both in the past and in the future. It is impor-
tant for defining a life trajectory, selecting more appropriate language
teaching methods, language policy, and making political and educa-
tional decisions.
Different features of multi-competence are called into action in different
situations and for various purposes, whether in language acquisition or in
social communication. Therefore selecting the variables according to the
aims and particular interest of a researcher and the demands of the situa-
tion will allow study of the same DLCs in their various facets with the
desired amount of detail. To this end, in each investigation of a DLC,
appropriate features in constitution, configuration and dynamics are to
be selected.
Which fields of research will benefit from the DLC approach? If the DLC
being researched is used by a large group of people, such knowledge could
be most useful in language acquisition. Studying cross-linguistic influ-
ences in typical DLCs would be welcome, as this work on CLI would be
more focused on particular populations. Multi-competence of a particular

Table 7.3 Features for investigation of multi-competence manifested in a DLC, with examples
Constitution (what/which) Configuration Dynamics

Including facts and descriptions Including numerical data and current patterns Alterations in the past and the future
Which languages? Comparative competence in each language Which languages used to be in a DLC for
Absolute competence in each Comparative time of use of each language this person?
Absolute time of use in each Comparative language modes for this DLC Any changes in the configuration?
Age of starting each language (possibly methods of learning, i.e.
tutored or untutored, intensive or sporadic, in the presence of
native speakers, etc.)
In which aspects (for example, education, material culture, One language may be used entirely or partially for education;
literature, popular songs) in a particular community are another may be well presented in the realm of everyday
languages of the community DLC transparent/transitive, materialities
interchangeable, and in which not? One language may be used only or mostly in the domain of
In which domains in a particular community are languages of the administration and education, another at home and for
community DLC interchangeable, and in which not, that is, which entertainment and fun, and the third with family and friends,
are strictly defined? or in the ethnic community in the case of minorities
Cultural-historical characteristics for a particular DLC, such as: There may be three different systems of numbering or colour How and in which direction were those
names in three languages, or the systems may coincide in aspects changed, if at all?
– colours in each language two languages but differ in the third (e.g. English twenty-one, Does a person use the colour system of
– numbering in each language Russian двадцать один, but German ein-und-zwanzig) one language of his DLC for all the
– grammars of each language In Russian and in French, plural and singular are used languages, or does he use the system
– pragmatic traditions in each language (e.g. forms of address in differentially to address older or younger, intimately or of the language he is using at the time?
face-to-face meeting and in writing) societally known people; in Hebrew and English the same How are pronouns used?
form of pronoun is used for all: you, ‫את‬
Present community characteristics Comparative role and status of each language of the Changes? News?
Community multilingual or not constellation in a given community
Community with separated language communities (bubbles) or
with blended culture?
Role and status of languages comprising the DLC in the global
hierarchy
population with a common DLC could be studied in depth. Exploring

multi-competence through DLC enables us to arrive at more informed
decisions regarding education and language teaching. It helps in dealing
with identification and representation of language users, and optimizing
practical responses to multilingualism – in terms of language policies and
educational measures.
When several DLCs are active in the same country or in neighbouring
countries, the exploration of their interactions will be informative both
for language contact studies and for the sociology of language. The
differences and similarities between the DLCs of the same region, and
consequently of specifics in their multi-competence would be helpful to
know: for example, the DLCs ‘in circulation’ in Israel, mentioned in
Section 7.3, include Arabic/Hebrew/English, Russian/Hebrew/English,
Romanian/German, Hebrew/English, and a number of other constella-
tions. It is enlightening to know which languages, which of their func-
tions, and which configurations coincide and differ, and the implications
of these differences and similarities.
As a first step in investigating language constellations, descriptive
studies are vital. There is a need to map the DLCs and to define
patterns that are common and rare, patterns that ‘work’ for career,
or integration into a community, or for mobility, or those DLCs that
might be rewarding in a niched style of life, somewhere in a small
local neighbourhood. Then correlation and comparison studies will be
useful too. Gradually communal DLCs will be studied synchronically
and also diachronically in terms of their dynamics and changes. With
the DLC, the possibility of collecting numerical data of all kinds
provides a good opportunity for research. Notably, a DLC approach
allows for both qualitative and quantitative research and is also suited
to complexity research methods.
With a DLC approach, we are approaching a way of measuring, at least
roughly, multilingualism in a community and a country, by marking and
assessing the ties between multilingual communities and the routes for
spreading vital information in a country. Counting and tracing the net-
works between DLCs of a country, or between countries, would show links
(in a graph where DLC are nodes, links would be drawn as edges, connect-
ing nodes) of each DLC inside a country, and globally.
In a country where there are many peripheral connections and fewer
central links, multilingualism would be very diverse and unstructured.
One of the possible challenges that such a kind of multilingualism may
create is the need to reiterate important things in many languages. In
the countries where there are many strong links, that is, the modes of
communities where different DLCs coincide, could it be suggested that
multilingualism is less diversified and, arguably, stronger? How does it
influence the vitality of languages in a certain context if their functions are
reiterated in a DLC?

160 LARISSA ARONIN
When acquiring a certain configuration a system becomes stable; other

configurations destabilize it. Assessment of DLCs or their configurations as
appropriate or successful for fulfilling numerous functions is also possible:
cognition, social coherence, career development, pursuing personal goals,
school aims may all be evaluated with the use of the DLC approach. Such an
assessment is possible, not in exact terms, but according to research aims.
We can ask which particular configurations in which particular environ-
ments make the holder of a DLC successful in these particular environments.
For another environment, e.g. moving into another country, or finding
oneself in the same country 20 years later, the previous DLC configuration
and contents most often do not fit, and are not stable in the new situation.
Finally, but also importantly, we can see from the above that the DLC
approach stimulates more questions on multi-competence, including the
following:
– Do all DLCs function in the same way in terms of the time devoted to
each language, domains and situations of use of L1, L2 and L3? Or do the
ways single languages in a constellation are employed and interact
(configuration patterns) differ from DLC to DLC?
– What are the cognitive and social implications of using various configu-
ration patterns?
7.5 Conclusion
Multi-competence is a highly appreciated perspective in language

acquisition and use which originated in the framework of SLA research,
and is now proving increasingly valid for circumstances of
multilingualism.
This chapter attempts to demonstrate how the Dominant Language
Constellation concept is related to the multi-competence perspective.
The associated concepts of multilinguality and language repertoire are
discussed with regard to their similarity and distinctness and their
relationship with multi-competence and DLC. Different individuals,
whether they have the same or different DLCs, each possess a unique
multilinguality. All bi- and multilinguals enjoy the faculty of multi-com-
petence, but as with other human characteristics, this is manifested in its
own way with each individual via his/her own DLC. Multi-competence as
an individual characteristic translates into the feature of a multilingual
society and is involved in language practices in a community. DLC is a
framework for situating multi-competence and also serves as an interface
between the individual and the communal.
The investigations with the help of a DLC can be carried out on at least
two planes: (1) sociolinguistic, in zooming into how, in which proportions
and for which results and implications constellations of languages are

used in society in communities; (2) in the realm of language acquisition,

how multi-competence emerges to be what it is, in the framework of a set
of particular languages.
Both as an abstract model and as the lens through which a particular real
situation of multi-competence is perceived, DLC carries out the functions
of systematizing, organizing, specifying and collecting the data regarding
multilinguals: how they deal with multiple languages in parallel, and how
multilingualism occurs in real places and communities. In addition, there
is the opportunity to assess and approximate the measuring of multilin-
gualism’s facets in a society. With the help of a DLC approach, a more exact
and unified picture could emerge, rather than just description and theore-
tical argumentation.
References
Aronin, L. 2006. ‘Dominant language constellations: an approach to multi-

lingualism studies’, in M. Ó Laoire (ed.), Multilingualism in Educational
Settings, pp. 140–159. Hohengehren: Schneider Publications.
Aronin, L. Fishman, J., Singleton, D. and Ó Laoire, M. 2013. ‘Introduction.
Current multilingualism: a new linguistic dispensation’, in D. Singleton
et al. (eds), pp. 3–23.
Aronin, L. and Jessner, U. 2014. ‘Methodology in bi- and multilingual
studies: from simplification to complexity’, in R. M. Manchón (ed.),
Research Methods and Approaches in Applied Linguistics: Looking Back and
Moving Forward, vol. 27 of AILA Review.
Aronin, L. and Ó Laoire, M. 2001. Exploring multilingualism in cultural
context: towards a notion of multilinguality. Paper presented at the
Third International Conference on Third Language Acquisition.
September, Leeuwarden, the Netherlands.
Aronin, L. and Ó Laoire, M. 2004. ‘Exploring multilingualism in cultural
contexts: towards a notion of multilinguality’, in C. Hoffmann and
J. Ytsma (eds), Trilingualism in Family, School and Community, pp. 11–29.
Aronin, L. and Ó Laoire, M. 2005. Thinking of multilinguality – ‘my self ‘or
‘my various selves’? An exploration of the identity of multilinguals. Paper
presented at Fourth International Conference on Multilingualism and
Multilingual Acquisition, September 8–10, Fribourg, Switzerland.
Aronin, L. and Singleton, D. 2008. ‘Multilingualism as a new linguistic
dispensation’, International Journal of Multilingualism 5 (1), 1–16.
Aronin, L. and Singleton, D. 2012. Multilingualism. Amsterdam: John
Benjamins.
Bois, M. 1993. Rudolf Noureev. Paris: Editions Plume. Online version,
accessed 17 February 2014: www.rudolfnureyev.org/reminiscences/mar
io_bua.html

162 LARISSA ARONIN
Canagarajah, S. 2007. ‘Lingua Franca English, multilingual communities

and language acquisition’, Modern Language Journal 91 (5), 921–937.
Caruana, S. and Lasagabaster, D. 2013. ‘Faraway, so close: trilingualism in
the Basque Autonomous Community and Malta from a social-educa-
tional perspective,’ in D. Singleton, J. A. Fishman, L. Aronin and
M. Ó Laoire (eds), Current Multilingualism: A New Linguistic Dispensation,
pp. 45–68. Berlin: Mouton De Gruyter.
tence’, Second Language Research 7 (2), 103–117.
557–591.
Cook, V. J. 2009. ‘Multilingual Universal Grammar as the norm’, in
Y.-K. I. Leung, (ed.), Third Language Acquisition and Universal Grammar,
pp. 55–70. Bristol: Multilingual Matters.
Cook, V. J. 2010. ‘The relationship between first and second language
acquisition revisited’, in E. Macaro (ed.), The Continuum Companion to
Second Language Acquisition, pp. 137–157. Continuum.
Cook, V. J. 2013a. ‘Multi-competence’, in C. A. Chapelle (ed.), The
Encyclopedia of Applied Linguistics, pp. 3768–3774. New York: Wiley-
Blackwell.
Cook, V. J. 2013b. ‘ELF: Central or atypical form of SLA?’, in D. Singleton et
al. (eds), pp. 27–44.
Coulmas, F. 2013. Sociolinguistics: The Study of Speakers’ Choices, second edi-
tion. Cambridge University Press.
De Saussure, F. 1916/1959. Cours de Linguistique Générale. Paris: Payot.
English translation by W. Baskin, 1959, Course in General Linguistics,
London: Peter Owen.
Fishman, J. A. 1998. ‘The new linguistic order’, Foreign Policy 113 (Winter),
26–40.
Gumperz, J. J. 1964. ‘Linguistic and social interaction in two communities’,
American Anthropologist 66 (6): 137–153.
Hamers, J. F. and Blanc, M. 1983. Bilingualité et Bilingualisme. Brussels:
Mardaga.
MacKenzie, I. 2012. ‘English as a lingua franca in Europe: bilingualism
and multi-competence’, International Journal of Multilingualism 9 (1),
83–100.
Popovic, R. 2009. ‘A long, adventurous road to (im)perfection’, in E. Todeva
and J. Cenoz (eds), pp. 33–49.
Pütz, M. 2004. ‘Linguistic repertoire / Sprachrepertoire’, in U. Ammon, N.
Dittmar, K. J. Mattheier and P. Trudgill (eds), An International Handbook of
the Science of Language and Society / Ein Internationales Handbuch zur
Wissenschaft von Sprache und Gesellschaft, second edition, pp. 226–231.
Berlin: Walter de Gruyter.
Schiffmann, H. F. 1996. Linguistic Culture and Language Policy. New York:
Routledge.

Singleton, D., Fishman, J. A., Aronin, L. and Ó Laoire, M. (eds) 2013. Current
Multilingualism: A New Linguistic Dispensation. Berlin: Mouton De Gruyter.
Todeva, E. and Cenoz, J. (eds) 2009. The Multiple Realities of Multilingualism:
Personal Narratives and Researchers’ Perspectives, pp. 33–49. Berlin: Mouton
De Gruyter.
Tonkin, H. 2009. ‘Where art and nature meet’, in E. Todeva and J. Cenoz
(eds), pp. 191–208.
Zipf, G. K. 1949/1965. Human Behavior and the Principle of Least Effort: An
Introduction to Human Ecology. New York and London: Hafner Publishing
Company.

8
Consequences of
multi-competence
for sociolinguistic
research
Li Wei
Sociolinguistics has been described as ‘the linguistics of choice’ (Coulmas

2013, p. 9), i.e. the study of the social conditions of who says what to
whom, how and why. Making appropriate choices is essential to successful
communication. The process of choice making, however, has not been a
core concern of formal linguistics, which, instead, is more interested in
delineating the general principles governing the structures of languages.
Psycholinguists are interested in choice making. But they see it primarily
or even solely as a cognitive process, and tend to be concerned more with
the neural networks or individual experiences that are responsible for
making linguistic choices. In this context, the notion of multi-competence
(MC) has important implications for the way we investigate the process of
choice making by language users. It brings together what has traditionally
been conceptualised separately as sociolinguistic and psycholinguistic
approaches, and broadens the scope of choice to include multimodality
and polysemiotic systems.
In this chapter I will focus on three of the central themes of sociolin-
guistics, namely language variation, language and social disadvantage, and
multilingual and intercultural communication, and present them from
an MC perspective. Whilst focusing on the implications of MC for socio-
linguistics, I will point out the connections between sociolinguistics and
psycholinguistics and other approaches. I will also discuss the implica-
tions of recent extensions of MC to the community, or collective MC. To
begin with, it would be useful to review three key concepts: communica-
tive competence, symbolic competence and MC. The first two are
frequently used in sociolinguistics. The connections between them and
MC serve as a good starting point for a discussion of the consequences of
MC for sociolinguistics.

Consequences of multi-competence for sociolinguistic research 165
8.1 Communicative competence, symbolic competence

and multi-competence
Communicative competence (CC) is central to the sociolinguistic enter-

prise, which is built on the premise that language is a social practice. In
addition to knowing how to form a grammatically correct sentence, one
needs to know, in order to communicate successfully through language,
the social roles and status of the interlocutors, the background knowledge
interlocutors share, and what interlocutors are trying to accomplish in
their interaction. This kind of knowledge is captured in the notion of CC,
coined by the anthropological linguist Dell Hymes (1966) in reaction to
the perceived inadequacy of Chomsky’s (1965) distinction between compe-
tence and performance. Hymes defines competence with the common sense
meaning of capability located in individual persons, not in the abstract
language system: ‘It cannot be assumed that the formal possibilities of a
system and individual knowledge are identical . . . I should take competence
as the most general term for the capabilities of a person’ (1972, p. 282).
Chomsky later (1980) talked about pragmatic competence, which involves
‘knowledge of conditions and manner of appropriate use, in conformity
with various purposes’ (Chomsky 1980, p. 224). Working with an ethno-
graphic framework, Hymes used the word SPEAKING as an acronym for
the various factors that a participant in a social interaction needs to
consider in order to communicate with others:
S The Setting and Scene of the speech event.

P The Participants of the event in various combinations of social
relationship and status.
E The Ends, or conventionally recognised and expected outcomes of an
exchange as well as personal objectives that a participant seeks to
accomplish on a particular occasion.
A The Act sequence of what is said and how.
K The Key, or tone, manner and spirit in which a particular message is
conveyed.
I The Instrumentalities, or the choice of channel and medium of
communication.
N: The Norms of interaction and interpretation for the occasion.
G: The Genre, such as telephone conversation, lecture, sermon, editorial,
riddle, poem, etc.
Hymes reminded us that speaking is a highly complex social activity,

involving knowledge and skills that are way beyond producing grammati-
cally correct sentences.
CC has had a fundamental impact on language teaching and learning
where the notion of social or situational appropriateness has become a
key objective (Cook 2007). Learning a language, whether first or second

166 LI WEI
language, entails developing what Hymes called ‘a general theory of the

speaking appropriate in their community, which they employ, like other
forms of tacit cultural knowledge (competence), in conducting and inter-
preting social life’ (1972, p. 279).
Influential as Hymes’s CC has been as a theoretical concept, critics
have raised the issue of the validity of the parallelism that Hymes’s for-
mulation seems to assume between grammaticality in Chomsky’s notion
of linguistic competence and appropriateness in his own. To limit
communicative competence to a rather mechanical correlation between
language and situation would be almost the same as stimulus and response
(see a fuller discussion in Cazden 2011). To be absolutely accurate, Hymes
does not define appropriateness as only the ability to respond in a
pre-existing context. In fact, he emphasised in many of his works the
importance of human ability to create contexts through language, or
what he called the emergent, creative, non-situationally determined use
of stylistic knowledge. Nevertheless, the notion of appropriateness raises
the question of whose social norm it should be judged by. What is con-
sidered appropriate by one group of speakers in a specific context may
be regarded as totally inappropriate by another group. As sociolinguists
have shown, the increasingly multilingual and multicultural nature of
global exchanges in today’s world means that one cannot assume that
the interlocutors in everyday social interaction all agree on what the
purpose of the exchange is and what constitutes a culturally appropriate
topic and style of conversation, and all have equal speaking rights and
opportunities (Johnstone 1996; Blommaert 2005; Coupland 2007). As
Rampton (1998, 1999) argues, language users have to mediate complex
encounters among interlocutors with different language capacities and
cultural imaginations, different social and political memories, and differ-
ent understandings of the social reality they live in. In order to get heard,
they need to exercise symbolic power using multiple semiotic resources.
It is within this context that Claire Kramsch (Kramsch 2006; Kramsch
and Whiteside 2008; Kramsch 2009) proposed the notion of symbolic
competence (SC):
Social actors in multilingual settings seem to activate more than a
communicative competence that would enable them to communicate
accurately, effectively, and appropriately with one another. They seem
to display a particularly acute ability to play with various linguistic codes
and with the various spatial and temporal resonances of these codes.
(Kramsch and Whiteside 2008, p. 664)
SC is defined as:
the ability not only to approximate or appropriate for oneself someone
else’s language, but to shape the very context in which the language is
learned and used. Such an ability is reminiscent of Bourdieu’s notion of
sens pratique, exercised by a habitus that structures the very field it is

structured by in a quest for symbolic survival (Bourdieu 1997/2000,

p. 150). (Kramsch and Whiteside 2008, p. 664)
Working with multilingual data from what she calls an ecological perspec-
tive, Kramsch argues that ‘a multilingual sens pratique . . . multiplies the
possibilities of meaning offered by the various codes in presence’. She
further points out that:
In today’s global and migratory world, distinction might not come so
much from the ownership of one social or linguistic patrimony (e.g.
Mexican or Chinese culture, English language) as much as it comes from
the ability to play a game of distinction on the margins of established
patrimonies. (Kramsch and Whiteside 2008, p. 664)
Because it depends on the other players in the game, Kramsch talks about
a ‘distributed’ SC that operates in four different ways:
(i) Subjectivity or subject-positioning. Different languages position their

speakers in different symbolic spaces. Subject positioning has to do
less with the calculations of rational actors than with multilinguals’
heightened awareness of the embodied nature of language and the
sedimented emotions associated with the use of a given language.
(ii) Historicity or an understanding of the cultural memories evoked by symbolic
systems. Any utterance or turn-at-talk can become a lieu de mémoire (or
‘place of memory’), formed by the sedimented representations of a
people. Whether these representations are accurate or not, histori-
cally attested or only imagined, they are actually remembered by
individual members and serve as valid historical models. SC is the
ability to perform and construct various historicities in dialogue with
others.
(iii) Performativity or the capacity to perform and create alternative realities.
Within an ecological perspective of human exchanges, utterances
not only perform some role or meaning, but they bring about that
which they utter, that is, they are performatives. Multilingual envir-
onments can elicit complex relationships between speech acts and
their perlocutionary effects. SC is thus the capacity to use the various
codes to create alternative realities and reframe the balance of sym-
bolic power.
(iv) Reframing as a means of changing the context as well as human thought and
action. SC is seen here as the ability to shape the multilingual game in
which one invests – the ability to manipulate the conventional cate-
gories and societal norms of truthfulness, legitimacy, seriousness,
originality.
Kramsch points out that SC is not reserved to multilingual actors in

multilingual encounters. Gumperz (1982) found that the meanings of
utterances in exchanges between monolingual speakers of English also

168 LI WEI
lie not only in the way participants orient themselves to the ongoing
exchange, but in the way they implicitly ventriloquise or even parody
prior utterances and thereby create affordances in ways that are favour-
able to them. Multilingual encounters increase the contact surfaces among
symbolic systems and thus the potential for creating multiple meanings
and identities. SC moved our attention away from appropriateness to
appropriation and the ability to actively manipulate and shape one’s
environment on multiple scales of time and space. In Kramsch’s words,
SC adds ‘a qualitative metalayer to all the uses of language studied by
applied linguists, one that makes language variation, choice, and style
central to the language learning enterprise’ (Kramsch and Whiteside
2008, p. 667).
Like CC and SC, MC also emphasises the individual capabilities of the
language user rather than systemic potentials of language. However, it is
Premise 3 of the concept of MC (see Chapter 1, this volume) – that it affects
the whole mind, i.e. all language and cognitive systems rather than
language alone – that particularly attracts me. Communication involves
multiple cognitive domains including memory, visual–spatial ability,
executive function and language. To me, the ability to use multisensory
and multimodal cues, including but not limited to verbal cues, is an
essential component of MC. To give a quick example, the heart sign has
become an iconic element in expressions of affection worldwide and has
been conventionalised in signs such as I ♥ London and I ♥ NY, where an
image of a heart, traditionally understood as a noun, is used in the place in
the linguistic construction that is usually occupied by a verb. Yet when the
expression is ‘read out’, most people would say ‘I love London’ and ‘I love
New York’ rather than ‘I heart London’ and ‘I heart New York’, so changing
the grammatical status of the word to which the image is linked. And
when people do say ‘heart’, as the author Lindsey Kelk does in her series
of popular books, it is for a special, often humorous, effect. During the
2008 US presidential election, there was a sign created in Ireland and
widely used by pro-Obama Irish and Irish American people where the
heart symbol was replaced by an image of a shamrock, and Obama’s
name was spelled in a mock Irish fashion O’Bama. Practically everybody
understood it as ‘I love Obama’, reinscribing the shamrock sign with the
meaning that is conventionally symbolised by the heart sign to mean
‘love’. Everyday communication in today’s world is full of examples like
this, simultaneously involving traditional linguistic signs (letters) and
images, emoticons and pictures. I have called it ‘translanguaging’ – the
dynamic process whereby language users mediate complex social and
cognitive activities through strategic employment of multiple semiotic
resources to act, to know and to be, that transcends conventional linguistic
systems (e.g. Li Wei 2011; Garcia and Li Wei 2014). Research evidence
shows that children, even infants, have no problem using their multiple
semiotic resources to interpret different forms of symbolic references

(Namy and Waxman 1998; Plester et al. 2011). Human beings have a
natural ‘translanguaging instinct’. It has also shown that multilingual
language users, especially those who live in communities where code-
switching is the social norm, are particularly good at translanguaging
(Kharkhurin and Li Wei 2015; Kharkhurin, Chapter 20, this volume).
In recent articles, Cook (2013; Chapter 1, this volume) has sought to
extend the concept of MC to include ‘the total system for all languages
(L1, L2, Ln) in a single mind or community and their inter-relationships’
(this volume, p. 7), and not just individual language users. This extension
connects well with the sociolinguistic agenda where the interests have
always been on speakers-in-community rather than individuals. As
Coulmas argues, ‘Individual act of choice does not reveal the social nature
of language. That only becomes apparent if we can show how individual
choices add up to form collective choices’ (2013, p. 11). Nevertheless,
collective MC can be a contentious issue. I will explore this further in
Section 8.5 of this chapter.
8.2 Linguistic variation as multi-competence
The metaphor that language is a living organism has been over- and
mis-used. Language does not have life of its own; it is the language user
who gives language life, changes it, and abandons it when it is no longer
needed. So-called language variation and change should in fact be how
language users vary and change their linguistic practices across different
contexts and times. Yet in existing sociolinguistic studies of language
variation and change, researchers seem to have pursued two separate
paths, largely following Labov’s approach (Labov 1994, 2001, 2010), of
investigating internally and externally motivated variation and change.
Internally motivated variations and changes are those that can be traced
to structural considerations in a language, whilst externally motivated
variations and changes are triggered and guided by social factors
(Hickey 2012). There is, however, considerable debate to as whether a
clear distinction can, or should, be drawn between internal and external
motivations. As Milroy (1992, p. 16) points out, the question is essentially
one which considers the ‘behaviour of speakers’ versus the ‘properties of
languages’. Let’s take speech accommodation as an example. It is well
known that speakers adapt their speech to accommodate interlocutors
and situational needs, e.g. talking to one’s own children versus talking to
doctors or accountants in a professional context, or talking to different
social groups (according to gender, ethnicity, socioeconomic class, place of
origin, language, etc.) about different topics (sports, politics, arts, holidays,
etc.). But the style shifting, or speech accommodation, needs to be done
within the structural constraints of the language or languages being used.
Speakers are usually aware of what is permissible and what is not within

170 LI WEI
the structural parameters. The permissible features will also need to be

structurally salient and marked in order that the speaker and listener can
notice the difference. Furthermore, the features are socially and socio-
linguistically significant because they carry specific meanings in specific
situational contexts and for specific speaker groups, such as glottal stops
in Tyneside English or the use of ain’t as a general negative indicator in
African American Vernacular English. The structural parameters are the
so-called internal factors for language variation and change. But how the
structural features became socially and sociolinguistically significant
would have to be traced to the speakers’ reactions to the speech of differ-
ent social groups in different contexts, which are external factors. The
structural parameters can rule out the choice of features that are alien to a
particular language. But, amongst the permissible choices, which feature
carries what social meaning derives from the speakers’ reactions to them,
which over time become community norms. My main point here is that a
simple binary division of linguistic variation and change into ‘internal’
and ‘external’ does not do justice to the complex and intricate relationship
between how speakers act linguistically in their community and the
postulated abstract level of structure which is taken to provide the basis
for speakers’ behaviour.
It is commonplace for sociolinguists to state that language use is
characterised by continual variation. From the MC perspective, it would
mean that speakers construct systemic knowledge of the language they
are exposed to both in terms of structural patterns including structural
variations and in terms of the social meanings of the structures. Moreover,
they continuously reanalyse and reconstruct their knowledge according
to the behavioural norms of the community as they begin to participate
in groupings of the society they are entering and when they become
integrated into this society. A competent language user would be able to
adapt themselves to specific communicative context. Seen in this light,
language variation and change would be an integral component of the
language user’s multi-competence.
8.3 Multi-competence and social disadvantage
Language variation and change as discussed in Section 8.2 presupposes

optionality and choice. Whilst optionality is a property or capacity of
the linguistic structure, choice is a property or capacity of the language
user. A fully developed language would normally have a full range of
optionality, in order to function for different stylistic and social pur-
poses. Endangered languages often lose their optionality as their usage
is confined to restricted contexts. Yet, given the linguistic options
available, how choices are made depends on individual language users
and the social circumstances they are in. It is important to remind

ourselves that not all options may be available equally to all language
users. This inequality has little to do with the linguistic system itself,
but is a social construct. In other words, whilst a language system may
contain various options, not all its users can access these options
equally because of cognitive, physical, educational, economic, socio-
cultural or political constraints. It is therefore important to investigate
and understand not only how choices are made, but also what choices
are available to which group of language users and why not all choices
are equally available to all.
Back in the 1970s, the sociologist Basil Bernstein argued that there was
a correlation between social class and the use of the so-called ‘elaborated
code’ or the ‘restricted code’. ‘Elaborated code’ is a way of using language
that spells out everything; it is structurally more complete and full of
details, so that people overhearing a conversation would be able to under-
stand it. ‘Restricted code’, on the other hand, is shorter, condensed and
requires background information and prior knowledge. It would therefore
be identifiable as an insiders’ code. Bernstein was at pains to stress,
Clearly one code is not better than another; each possesses its own aes-
thetic, its own possibilities. Society, however, may place different values
on the orders of experience elicited, maintained and progressively
strengthened through the different coding systems. (Bernstein 1971, p. 135)
As an educator, Bernstein was interested in accounting for the rela-

tively poor performance of working-class students in language-based
subjects, when they were achieving scores as high as their middle-class
counterparts on mathematical topics. Whilst the restricted code works
better than the elaborated code for situations in which there is a great
deal of shared and taken-for-granted knowledge in the group of speakers,
the language used in schools is mainly that of the elaborated code. It
favours the middle class who, being more geographically, socially and
culturally mobile, has access to both the restricted codes and elaborated
codes, whereas working-class individuals tend to have access only to
restricted codes.
Bernstein’s thesis has also been misinterpreted or misunderstood. He
was never interested in the structural configurations of the codes them-
selves and argued very explicitly that one set of codes was not necessarily
better than another. Indeed, he pointed out that the ‘restricted code’ is
economical and rich, conveying a vast amount of meaning with a few
words, each of which has a complex set of connotations and acts like an
index, pointing the hearer to a lot more information which remains
unsaid. Bernstein’s main objective was to highlight the social factors that
determine the access to elaborated or restricted codes. These factors,
according to Bernstein, are sociocultural values and the socialising agen-
cies. In a society which values individuality, elaborated codes are privi-
leged, and in a society which values collective identities, restricted codes

172 LI WEI
tend to be more prominent. And when the socialising agencies (e.g. family,
peer group, school, work) are well defined and structured, one would find
restricted codes; conversely, where the agencies are malleable, elaborated
codes are found. As Bernstein says, ‘The orientation towards these codes
may be governed entirely by the form of the social relation, or more
generally by the quality of the social structure’ (1971, p. 135).
In societies where multiple different languages exist, beliefs and
attitudes persist that bilingualism disadvantages rather than advan-
tages one’s intellectual development. The early research on bilingual-
ism and cognition tended to confirm such negative views, finding that
monolinguals were superior to bilinguals on intelligence tests. One of
the most widely cited studies was done by Saer (1923) who studied
1,400 Welsh–English bilingual children between the ages of seven
and 14 in five rural and two urban areas of Wales. A 10-point difference
in IQ was found between the bilinguals and the monolingual English
speakers from rural backgrounds. From this Saer concluded that bilin-
guals were mentally confused and at a disadvantage in intelligence
compared with monolinguals. It was further suggested, with a follow-
up study of university students, that ‘the difference in mental ability as
revealed by intelligence tests is of a permanent nature since it persists
in students throughout their university career’ (1923, p. 53).
Controversies regarding the early versions of IQ tests and the definition
and measurement of intelligence aside, there were a number of problems
with Saer’s study and its conclusions. First, it appeared to be only in the
rural areas that the correlation between bilingualism and lower IQ held. In
urban areas monolinguals and bilinguals were virtually the same; in fact
the average IQ for urban Welsh–English bilingual children in Saer’s study
was 100, whereas for monolingual English-speaking children it was 99.
The urban bilingual children had more contact with English both before
beginning school and outside school hours than did the rural bilinguals.
Thus the depressed scores of the rural population were probably more a
reflection of lack of opportunity and contexts to use English and were not
necessarily indicative of any socio-psychological problems (see Peal and
Lambert 1962 for rather different results, though the selection of speakers
in their study was not without bias).
More important, however, is the issue of statistical inference in this and
other studies of a similar type. Correlations do not allow us to infer cause
and effect relationships, particularly when other variables – such as rural
versus urban differences – may be mediating factors. Another major factor
is the language in which such tests were administered, particularly tests
of verbal intelligence. Many such studies measured bilinguals only in the
second or non-dominant language.
At around the same time as Saer conducted studies on bilinguals’
intelligence, some well-known linguists expressed their doubts about
bilingual speakers’ linguistic competence. The following is Bloomfield’s

characterisation of a Menomini Indian man in the US who he believed

to have ‘deficient’ knowledge of Menomini and English:
White Thunder, a man around 40, speaks less English than Menomini, and
that is a strong indictment, for his Menomini is atrocious. His vocabulary
is small, his inflections are often barbarous, he constructs sentences of a
few threadbare models. He may be said to speak no language tolerably.
(Bloomfield 1927, p. 395)
This is one of the early statements of a view which became fashionable

in educational circles; namely, that it was possible for bilinguals not to
acquire full competence in any of the languages they spoke. Such an
individual was said to be ‘semilingual’ (Hansegard 1968). They were
believed to have linguistic deficits in six areas of language (see
Hansegard 1968; Skutnabb-Kangas 1981):
1 size of vocabulary;
2 correctness of language;
3 unconscious processing of language;
4 language creation;
5 mastery of the functions of language;
6 meanings and imagery.
It is significant that the term ‘semilingualism’ emerged in connection

with the study of language skills of people belonging to ethnic minority
groups. Research which provided evidence in support of the notion of
‘semilingualism’ was conducted in Scandinavia and North America and
was concerned with accounting for the educational outcomes of submer-
sion programmes where minority children were taught through the
medium of the majority language (Skutnabb-Kangas 1981). However,
these studies, like the ones conducted by Saer, had serious methodological
flaws and the conclusions reached by the researchers were misguided.
First, the educational tests which were used to measure language profi-
ciencies and to differentiate between people were insensitive to the
qualitative aspects of languages and to the great range of language compe-
tences. Language may be specific to a context; a person may be competent
in some contexts but not in others. Second, bilingual children are still in
the process of developing their languages. It is unfair to compare them to
some idealised adults. Their language skills change over time. Third, the
comparison with monolinguals is also unfair. It is important to distinguish
if bilinguals are ‘naturally’ qualitatively and quantitatively different from
monolinguals in their use of the two languages, i.e. as a function of being
bilingual. Fourth, if languages are relatively underdeveloped, the origins
may not be in bilingualism per se, but in the economic, political and social
conditions that evoke under-development. The disparaging and belittling
overtone of the term ‘semilingualism’ itself invokes expectations of under-
achievement in the bilingual speaker. Thus, rather than highlighting

174 LI WEI
the apparent ‘deficits’ of bilingual speakers, it is more important to

understand what conditions put certain bilinguals in a ‘semi’ state (for a
critical analysis of the notion of semilingualism, see Martin-Jones and
Romaine 1986).
One of the specific issues Bloomfield raised in his comments on the
language behaviour of members of the Menomini Indians in North
America was the frequent mixing of their own language and English. It
has been described as ‘verbal salad’, not particularly appealing but never-
theless harmless, or ‘garbage’ which is definitively worthless and vulgar.
Unfortunately, although switching and mixing of languages occurs in
practically all bilingual communities and all bilingual speakers’ speech,
it is stigmatised as an illegitimate mode of communication, even some-
times by the bilingual speakers themselves. Haugen (1977, p. 97), for
example, reports that a visitor from Norway made the following comment
on the speech of the Norwegians in the United States: ‘Strictly speaking, it
is no language whatever, but a gruesome mixture of Norwegian and
English, and often one does not know whether to take it humorously or
seriously.’ Gumperz (1982, pp. 62–63) reports that some bilingual speakers
who mixed languages regularly still believe such behaviour was ‘bad
manners’ or a sign of ‘lack of education or improper control of language’.
One of the Punjabi–English bilinguals Romaine interviewed said:
I’m guilty as well in the sense that we speak English more and more and
then what happens is that when you speak your own language you get
two or three English words in each sentence . . . but I think that’s wrong.
(Romaine 1995, p. 294)
In this regard, the ground-breaking studies by Bialystok and her

associates have proved that bilingual individuals consistently outperform
monolinguals on task-switching executive control tasks, suggesting that
bilingualism provides an advantage in the non-verbal, executive control
aspect of cognitive development (Bialystok and Barac 2012). Moreover,
bilingual children have been shown to have better developed phonological
and metalinguistic awareness, which is a necessary prerequisite in literacy
acquisition (Bialystok 2001; Bialystok and Luk 2007), and the bilingual
advantage seen in children may persist, and even strengthen, throughout
adulthood (Bialystok et al. 2009; Luk et al. 2011). Bialystok and her col-
leagues have argued that the constant management of two equally promi-
nent languages competing for attention recruits the use of executive
functions more often, and that this ability continually enhances functions
of executive control. This advantage may extend further and possibly help
to slow the natural decline of cognitive control brought about by the aging
process (Schweizer et al. 2010).
An interesting corollary has been provided by Kharkhurin who has
examined the link between multilingualism and creativity (Chapter 20,
this volume; Kharkhurin 2012). Through a series of experiments on

selective attention and language mediated concept activation, Kharkhurin

provides research evidence showing that multilingualism facilitates gen-
erative and innovative capacities of creative thinking. In a recent study
by Kharkhurin and Li Wei (2015), multilinguals who habitually switch
between different languages in their everyday social interaction have
more developed creativity as measured by the Torrance Creativity Tests.
Code-switching is a particularly interesting linguistic phenomenon in
the present consideration, as it appears to have a ‘semilingual’ surface
structure and is often assumed to be caused by deficits in language profi-
ciency or cognitive control. Yet these studies suggest that code-switching
is an indicator of multi-competence and can have positive and long-term
effects on the multilingual language users’ well-being. Other studies
have also shown that multilinguals who regularly use and switch between
their languages are more advanced in other aspects of social cognition
such as empathy and tolerance of ambiguity (Dewaele and Li 2012, 2013).
8.4 Multi-competence and multilingual and intercultural

communication
So far, the discussion of competence and multi-competence has centred

round language and cognition. What about ‘culture’? In the increasingly
globalised world today, should MC also include intercultural competence?
The question has been alluded to in Kramsch’s conceptualisation of
symbolic competence, discussed above and in Grosjean (2015), who
clearly sees the bilingual–bicultural as a specific group of bilinguals who
not only manoeuvre themselves between different languages but also
between different cultures. Byram (1997, pp. 48–53) proposed a model of
Intercultural Communicative Competence (ICC) in which an ‘intercultural
speaker’ mediates between different perspectives and cultures. The model
consists of the following parts.
– Attitudes (savoir être): curiosity and openness, readiness to suspend dis-

belief about other cultures and belief about one’s own. Other key words
are ‘cultural sensitivity’, ‘tolerance of ambiguity’, ‘respect of otherness’,
and ‘empathy’.
– Knowledge (savoir): knowledge of social groups and their products and
practices in one’s own and in one’s interlocutor’s country, and of the
general processes of societal and individual interaction. Knowledge of
a culture that goes beyond literature and well-known pieces of art
(so-called high culture, or Culture with a capital C), but also everyday
phenomena in relation to time, distance and nearness.
– Skills (I) of interpreting and relating (savoir comprendre): ability to inter-
pret a text or an event from another culture, to explain it and relate it to
texts or events from one’s own culture.

176 LI WEI
– Skills (II) of discovery and interaction (savoir apprendre/faire): ability to

acquire new knowledge of a culture and cultural practices and the
ability to operate knowledge, attitudes and skills under the constraints
of real-time communication (face-to-face encounters) and interaction.
– Critical cultural awareness (savoir s’engager): ability to perceive (see, hear
and feel) and to evaluate, critically and according to explicit criteria,
perspectives, practices and products in their own and other cultures and
countries.
Byram differentiates Intercultural Competence from Intercultural

Communicative Competence. The former refers to the ability to interact
in one’s own language with people from another culture and does not
necessarily require knowledge of the target language, while the latter is
about interacting with people from another culture in a second or foreign
language. ICC should also be differentiated from ‘multilingual compe-
tence’, which is sometimes confused with MC. Multilingual competence
is mostly used as a lay term referring to the ability to use several different
languages; it has no special theoretical value. A person who can speak
several languages may not have any significant knowledge of the cultures
the languages represent. In this regard, it is worth noting that the Council
of Europe, in its Common European Framework of Reference for
Languages where ICC is developed and promoted, makes a distinction
between plurilingualism and multilingualism. Whilst multilingualism is
about the coexistence of different languages, whether in the mind of
an individual or a society, and represents a diversification of languages
available, plurilingualism emphasises that a person’s languages and cul-
tures are not kept in strictly separated mental compartments, but that
the languages in one’s head will inter-relate and interact to contribute to
the growth of communication skills as a whole.
ICC raises the issue of the relationship between language and thought,
as speaking two languages may or may not entail thinking in two lan-
guages, and vice versa. There are, of course, opposing, extreme positions:
one is that thinking cannot take place without language; therefore how a
language is structured would impact on the way speakers of that language
think. This is the core of the ‘linguistic relativity’ position. The other
position, in contrast, would be that people think not in any particular
language but in a language of thought known as mentalese, a representa-
tion of concepts and propositions in the brain. This is apparently evident
among babies before they learn to speak, adults with aphasia, primates
who are capable of recognising kinships and others. Most of the arguments
are based on research evidence from monolingual speakers or cross-
linguistic comparisons. Recently, studies on bilingual speakers have
offered new insights into the relationship between language and thought,
or more precisely, the relationships between language, culture and
thought. Pavlenko (2006) provides an overview of the findings from

bilingualism research which have either direct or indirect bearings on the

debate in a number of areas, including colour, categorisation of objects
and substances, number, space, motion, time, emotions and personhood.
Bilinguals are shown to be able to draw on different linguistic and emo-
tional repertoires, stances and interpretations and to form distinctive
conceptualisations in different languages. Furthermore, Pavlenko (2011)
identifies seven processes of conceptual restructuring in the bilingual
mind, very much in the spirit of the MC premises. These seven pro-
cesses are:
– coexistence of L1 and L2 categories;

– the influence of L1 on L2 categories;
– convergence of L1 and L2 categories;
– restructuring in the direction of L2 categories;
– internationalisation of new categories absent in L1;
– the influence of L2 on L1;
– attribution of L1 categories.
These restructuring processes may be domain-specific and conditioned

by age of acquisition of L2, context of acquisition and length of expo-
sure, language proficiency and frequency of language use. Studies of
bilingualism as reviewed by Pavlenko help to discard the narrow search
for evidence for or against linguistic relativity and to engage in broad
explorations of thinking and speaking in two or more languages (2011,
p. 252)
8.5 Collective multi-competence
As mentioned above, Cook has in recent years sought to include ‘the

total system for all languages (L1, L2, Ln) in a single mind or community
and their inter-relationships’ in MC (Chapter 1, this volume, p. 7). This
has significant implications for sociolinguistic research. It is interesting
to observe, first of all, that communities in post-modern societies nowa-
days are often defined along language and ethnicity lines, rather than
the traditional, geographically based boundaries. For instance, families
of Scottish, Indian, Chinese, and Polish backgrounds may be living in
the same street but feel that they belong to different communities,
whilst members of the Scottish, Indian, Chinese and Polish commu-
nities may be living in different parts of the country or indeed the
world. Most of these communities would have several different lan-
guages being used by different generations of individuals, in different
contexts and for different purposes. The complex community history
and current configurations lead to differential status of languages, with
some languages having more social prestige than others. Spanish, for
example, is a high-prestige language in the Hispanic communities in

178 LI WEI
the US, but it is English that has the status of the official language of the
nation. However, different members of the community may feel differ-
ently with regard to the social significance of specific languages. First
generations of Indian immigrants in Britain may feel deeply attached to
their ethnic languages rather than English; their British-born children
and grandchildren, however, may feel that English is their primary
language. Such changes in attitude can contribute to changes in lan-
guage choice patterns across generations within a family or commu-
nity, which in turn can lead to language shift, i.e. collective adoption of
one language in place of another in the community. Just as the learning
and use of L2 would impact fundamentally on a language user’s L1, a
fact that has been highlighted in the conceptualisation of MC, the
adoption of a new language by a community collectively would change
the community fundamentally too. New practices, attitudes and ideol-
ogies will emerge. And a different identity, and most likely a different
set of values, of the community will be developed from the one that was
using another language previously.
The differential status of languages and intergenerational language
shift that characterise many multilingual communities raise the ques-
tion why different languages cannot coexist peacefully. Why should
there be such competition between languages? The answer must in part
be related to the wide-spread and deep-rooted monolingual ideology, or
as Baetens Beardsmore (2003) put it, ‘fears of bilingualism’, especially
the so-called ‘cultural fears’. Such fears motivate the perception and
belief that bilinguals must somehow undergo a form of disorientation
or acculturation brought about by cultures in conflict; they must have
difficulties in reconciling the two cultures embodied in the different
languages. T’sou (1985) strikingly puts it in his discussion of Anglo-
Chinese bilinguals in Hong Kong, the ‘cultural eunuch syndrome’.
Garmadi (1972, p. 319) uses equally strong terms to describe cultural
destabilisation brought about by bilingualism in developing nations,
where he talks of linguistic and cultural mutilation. However, the solu-
tions put forward in order to avoid problems of acculturation, alienation,
marginality, conflicts of identity and rootlessness are often the insistence
on one language only or one language at a time. To combat such fears and
the monolingual ideology, we need evidence to show that bilingualism
and multilingualism are important assets of a competent and confident
community or a nation.
Surveying the sociolinguistic situations in the European Union coun-
tries, Grin (1995, 1996, 2007) has argued that multilingualism plays a
key role in intercultural dialogue, social cohesion, economic prosperity,
lifelong learning, media and information technologies, as well as in the
EU’s external relations. In particular, a multilingual workforce is a distinct
advantage that would provide European companies and their countries
with a competitive edge and thus promote prosperity of the nation. From

the citizens’ perspective, multilingual competence increases employabil-

ity and allows them to choose from a larger number of job offers.
Moreover, multilingualism is shown to enhance creativity and innovation
of the individual as well as the nation, and student and worker mobility
across Europe and beyond. In sum, nations that have a positive attitude
towards multilingualism are nations with cultural vitality and at ease
with themselves. It is worth noting that in times of economic downturn
and crisis, there is often an increased amount of public hostility towards
multilingualism associated with migration. Such is the case in Britain,
where migrants are uniformly assumed to be lacking in English and need-
ing language support.
Multi-competence as a concept originally came out of second language
research, and has had a fundamental impact on the areas of second
language learning and bilingualism. The impact is most felt amongst
psycholinguists. This chapter has attempted to highlight the connections
with sociolinguistics and the consequences of adopting an MC perspective
for sociolinguistic research. In particular, we can reconceptualise linguis-
tic variation and change not simply as a property of the language system
but as a capacity of the language user. We can also address some of the
misconceptions of bilingualism and bilingual behaviour, and redirect
our analytic attention from a deficit model of bilingualism to a multi-
competence model. As it stands now, MC does not specifically address
the issue of culture and intercultural competence. This chapter has
shown that there are intrinsic links between the notions of Intercultural
Communicative Competence and MC. The recent extension of MC to
cover collective competence of the community should also open up new
venues of investigation, bringing psycholinguistics and sociolinguistics
closer together.
References
Baetens Beardsmore, H. 2003. ‘Who’s afraid of bilingualism?’, in

J.-M. Dewaele, A. Housen and Li Wei (eds), Bilingualism: Beyond Basic
Principles, pp. 10–27. Clevedon: Multilingual Matters.
Bernstein, B. 1971. Class, Codes and Control, vol. 1: Theoretical Studies Towards a
Sociology of Language. London: Routledge and Kegan Paul.
Bialystok, E. and Barac, R. 2012. ‘Bilingual effects on cognitive and linguis-
tic development: role of language, cultural background, and education’,
Child Development 83, 413–422.
Bialystok, E. and Luk, G. 2007. ‘Common and distinct cognitive bases for
reading in English–Cantonese bilinguals’, Applied Psycholinguistics 29,
269–289.

180 LI WEI
Bialystok, E., Luk, G. and Craik, F. 2009. ‘Cognitive control and lexical
access in younger and older bilinguals’, Journal of Experimental
Psychology: Learning, Memory, and Cognition 34 (4), 859–873.
Blommaert, J. 2005. Discourse. Cambridge University Press.
Bloomfield, L. 1927. ‘Literate and illiterate speech’, American Speech 2 (10),
432–439.
Bourdieu, P. 1997/2000. Pascalian Meditations, translated by R. Nice. Stanford
University Press.
Byram, M. 1997. Teaching and Assessing Intercultural Communicative
Competence. Clevedon: Multilingual Matters.
Cazden, C. B. 2011. ‘Dell Hymes’s construct of “Communicative
Competence” ’, Anthropology and Education Quarterly 42 (4), 364–369.
Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge: MIT Press.
Chomsky, N. 1980. Rules and Representations. Oxford: Basil Blackwell.
Cook, V. J. 2007. Second Language and Language Teaching, fourth edition.
London: Hodder Education.
Cook, V. J. 2013. ‘The language of the street’, Applied Linguistics Review 4 (1),
43–81.
Coulmas, F. 2013. Sociolinguistics: The Study of Speakers’ Choices, second edi-
tion. Cambridge University Press.
Coupland, N. 2007. Style: Language Variation and Identity. Cambridge
University Press.
Dewaele, J.-M. and Li Wei 2012. ‘Multilingualism, empathy and multi-
competence’, International Journal of Multilingualism 9 (4), 352–366.
Dewaele, J.-M. and Li Wei 2013. ‘Is multilingualism linked to a higher
tolerance of ambiguity?’ Bilingualism: Language and Cognition 16 (1),
231–240.
Education. Basingstoke: Palgrave Macmillan.
Garmadi, S. 1972. ‘Les problèmes du bilinguisme en Tunisie’, in
A. Abdel-Malek (ed.), Renaissance du Monde Arabe, pp. 309–322.
Gembloux: Duculot.
Grin, F. 1995. ‘The economics of language competence: a research project
of the Swiss National Science Foundation’, Journal of Multilingual and
Multicultural Development 16, 227–231.
Grin, F. 1996. ‘The economics of language: survey, assessment and pro-
spects’, International Journal of the Sociology of Language 121, 17–44.
Grin, F. 2007. ‘Economics and language policy’, in M. Hellinger and
A. Pauwels (eds), Handbook of Language and Communication: Diversity and
Change, pp. 271–297, vol. 9 of Handbook of Applied Linguistics. New York
and Berlin: Mouton De Gruyter.
Grosjean, F. 2015. ‘Bicultural bilinguals’, International Journal of Bilingualism
19 (5), 572–586.
Gumperz, J. J. 1982. Discourse Strategies. Cambridge University Press.

Hansegard, N. 1968. Tvatsprakighet eller Halvspra-kighet. Stockholm: Alders/

Banners.
Haugen, E. 1977. ‘Norm and deviation in bilingual communities’, in
P. A. Hornby (ed.), Bilingualism: Psychological, Social, and Educational
Implications, pp. 91–102. New York: Academic Press.
Hickey, R. 2012. ‘Internally and externally motivated language change’, in
J. M. Hernández-Compoy and J. C. Conde-Silvestre (eds), The Handbook of
Historical Sociolinguistics, pp. 401–421. Oxford: Wiley-Blackwell.
Hymes, D. H. 1966. ‘Two types of linguistic relativity’, in W. Bright (ed.),
Sociolinguistics, pp. 114–158. The Hague: Mouton.
Hymes, D. H. 1972. ‘On communicative competence’, in J. B. Pride and
J. Holmes, Sociolinguistics: Selected Readings. Harmondsworth: Penguin,
269–293.
Johnstone, B. 1996. The Linguistic Individual: Self-expression in Language and
Kharkhurin, A. V. 2012. Multilingualism and Creativity. Clevedon:
Kharkhurin, A. V. and Li Wei 2015. ‘The role of code-switching in bilingual
creativity’, International Journal of Bilingual Education and Bilingualism 18 (2),
153–169.
Kramsch, C. 2006. ‘From communicative competence to symbolic compe-
tence’, Modern Language Journal 90, 249–252.
Kramsch, C. 2009. The Multilingual Subject: What Language Learners Say about
their Experience and Why it Matters. Oxford University Press.
Kramsch, C. and Whiteside, A. 2008. ‘Language ecology in multilingual
settings: towards a theory of symbolic competence’, Applied Linguistics 29,
645–671.
Labov, W. 1994. Principles of Linguistic Change, vol. 1: Internal Factors. Oxford:
Blackwell.
Labov, W. 2001. Principles of Linguistic Change, vol. II: Social Factors. Oxford:
Wiley-Blackwell.
Labov, W. 2010. Principles of Linguistic Change, vol. III: Cognitive and Cultural
Factors. Oxford: Wiley-Blackwell.
Li Wei. 2011. ‘Moment analysis and translanguaging space: discursive
construction of identities by multilingual Chinese youth in Britain’,
Journal of Pragmatics 43 (5), 1222–1235.
Luk, G., Bialystok, E., Craik, F. I. M. and Grady, C. L. 2011. ‘Lifelong bilin-
gualism maintains white matter integrity in older adults’, The Journal of
Neuroscience 31, 16808–16813.
Martin-Jones, M. and Romaine, S. 1986. ‘Semilingualism: a half-baked
theory of communicative competence’, Applied Linguistics 7 (1),
26–38.
Milroy, J. 1992. Linguistic Variation and Change: On the Historical Sociolinguistics
of English. Oxford: Blackwell.

182 LI WEI
Namy, L. L. and Waxman, S. R. 1998. ‘Words and gestures: infants’ inter-

pretations of different forms of symbolic reference’, Child Development
69 (2), 295–308.
Pavlenko, A. (ed.) 2006. Bilingual Minds: Emotional Experience, Expression and
Representation. Clevedon: Multilingual Matters.
Pavlenko, A. (ed.) 2011. Thinking and Speaking in Two Languages. Clevedon:
gence’, Psychological Monographs: General and Applied 76 (27), 1–23.
Plester, B., Lerkkanen, M.-K., Linjama, L. J., Rasku-Puttonen, H. and
Littleton, K. 2011. ‘Finnish and UK English pre-teen children’s text mes-
sage language and its relationship with their literacy skills’, Journal of
Computer Assisted Learning 27 (1), 37–48.
Rampton, B. 1998. ‘Language crossing and the redefinition of reality’, in
P. Auer (ed.), Code-switching in Conversation, pp. 290–320. London:
Routledge.
Rampton, B. 1999. ‘Styling the other: introduction’, Journal of Sociolinguistics
3 (4), 421–427.
Romaine, S. 1995. Bilingualism, second edition. Malden, MA: Blackwell.
Saer, D. J. 1923. ‘An inquiry into the effect of bilingualism upon the
intelligence of young children’, Journal of Experimental Psychology. Part 1:
6, 232–240. Part 2: 6, 266–274.
Schweizer, T. A., Ware, J., Fischer, C. E., Craik, F. I. M. and Bialystok, E.
2010. ‘Bilingualism as a contributor to cognitive reserve: evidence from
brain atrophy in Alzheimer’s disease’, SciVerse ScienceDirect 48, 991–996.
Skutnabb-Kangas, T. 1981. Bilingualism or not: the Education of Minorities.
T’sou, B. 1985. ‘Chinese and the cultural eunuch syndrome’, in R. Lord and
B. T’sou (eds), The Language Bomb, pp. 15–19. Hong Kong: Longman.

9
A usage-based account
of multi-competence
Joan Kelly Hall
9.1 Introduction
One of the more significant outcomes of the processes of globalization is

the ever-increasing linguistic diversity of communicative practices invol-
ving individuals, communities, businesses and governments of different
states, nations and nation-states. As the range and number of languages
used in these practices have become more diverse and complex, so have
the linguistic resources individuals need to participate in them. Across
communities around the world, bilingualism and multilingualism are now
the standard rather than the exception.
By and large the research efforts of second language acquisition (SLA), a
discipline devoted to understanding how individuals acquire languages in
addition to their first language, have been hampered in its efforts to
understand how individuals develop the competence they need to partici-
pate in such bilingual and multilingual practices. The crux of the problem,
critics have long argued, lies in the field’s continued reliance on a view of
language competence as a fixed homogeneous system that takes the idea-
lized monolingual speaker as its norm (Blackledge 2005; Liddicoat and
Crichton 2008; Makoni and Pennycook 2007; Ortega 2013). This “mono-
lingual bias” (Ortega 2013) has its roots in formal linguistics whose view of
language competence, defined as an internal system of abstract structures,
was based on “an ideal speaker–listener, in a completely homogeneous
speech-community” (Chomsky 1965, p. 3).
Arguing for the need of a concept of competence that more
adequately captured what bilinguals and multilinguals know, early in
the 1990s, Vivian Cook (1991, 1992) proposed the term multi-
competence. Defined as a “compound state of mind with two gram-
mars” (Cook 1991, p. 112), multi-competence became a springboard for
research on bilingual and multilingual language knowledge and
gave the field a useful lens for viewing L2 learners as successful

184 J O A N K E L LY H A L L
multi-competent language users rather than as unsuccessful or defi-

cient native speakers.
While noting the significant role the concept of multi-competence
played in helping to transform the field’s views of L2 learners, Hall,
Cheng and Carlson (2006) argued that in retaining a view of L1 and L2
language knowledge as discrete systems and assuming homogeneity
of language knowledge across speakers and contexts, the concept was
less useful in transforming understandings of language knowledge. To
remedy this, they argued for a usage-based view of multi-competence.
The term usage-based is commonly used as an umbrella term to cover
research from cognitive science disciplines such as developmental
psychology, neurolinguistics and cognitive and functional linguistics
concerned with explaining how language knowledge emerges from
the interaction between cognition and language use (e.g. Bates et al.
1998; Bybee 2010; Elman 1999; Ibbotson 2013; Langacker 1987; Lee
et al. 2009; Levinson 2006; MacWhinney 2008; Tomasello 2006).
In this chapter I elaborate on the usage-based account of multi-
competence presented in Hall, Cheng and Carlson (2006). Here, I am
broadening the use of the term to include linguistic ethnographic
approaches to the study of language use and development (e.g.
Blommaert and Backus 2011; Collins 2014; Gumperz 1982; Halliday
1975, 1978; Hanks 1996; Hymes 1974; Ochs and Schieffelin 2008). While
the research foci and methodologies of the cognitive-oriented approaches
differ from those of the social-oriented approaches, their views of
language knowledge as driven by use make them theoretical allies.
Together, their insights and empirical findings converge to form a compel-
ling view of the multi-competent individual.
9.2 A usage-based account of multi-competence
In contrast to a formal view of language knowledge as unchanging, invar-

iant forms comprising systems of abstract structures, a usage-based per-
spective of multi-competence considers language knowledge to be
fundamentally dynamic and provisional, formed as “a massive collection
of heterogeneous constructions” (Bybee and Hopper 2001b, p. 3, emphasis in
the original). These constructions are grounded in and emergent from
concrete uses in culturally framed communicative activities, which, in
turn, are tied to specific sociocultural communities of practice.
Accordingly, rather than a prerequisite to individual language use, lan-
guage knowledge develops from our experiences with language as we go
about doing “the communicative work humans do” (p. 48), with the regu-
larities of individuals’ language knowledge emerging from their “lifetime
analysis of the distributional characteristics of the language input and
their usage” (Ellis 2006, p. 9).

A usage-based account of multi-competence 185
Constructions are conventionalized form–meaning combinations that

vary in size and degree of analytic specificity and, so, include single words,
collocations of words and other groupings of units comprising idioms and
routine formulas and schemas as well as fully abstract constructions (Boyd
and Goldberg 2009). The dynamic repertoires of constructions that indivi-
duals develop from their real-world experiences are cognitive in that they
are represented in individual minds as relatively automatized, context-
sensitive collections of forms and patterns. They are, at the same time,
social in that their development is an ever-evolving process as individuals
navigate their way through their multilingual contexts of action over their
lifespans (Halliday 1975; Hymes 1972; Ochs 1988; Vygotsky 1978).
In this view, then, competence and performance are not separate phe-
nomena. Rather, they are interdependent, “with competence being the
integrated sum of prior usage and performance being its dynamic contex-
tualised activation” (Ellis 2006, p. 100).
Any appearance of stability in individuals’ language knowledge is a
matter of stability in their linguistic experiences, with similar construc-
tions across users a reflection of “historically popular solutions to similar
communicative and coordination problems” (Ibbotson 2012, p. 124),
rather than of an innate universal system of language.
9.2.1 Cognitive-oriented account

Research from the various disciplines within cognitive science, including
developmental psychology, neurolinguistics and cognitive and functional
linguistics, provides a compelling account for the various types of cogni-
tive processes that are entailed in language learning and the specific ways
they interact with properties of linguistic input to give shape to, regularize
and automatize individual cognitive representations of language. The
internal mechanisms consist of three broad collections of cognitive capa-
cities. A first set includes innate domain-general cognitive constraints,
such as architectural constraints, which refer to the inherent structuring
of the brain that makes possible the processing of information, and chron-
otopic constraints, defined in terms of the timing of events in the devel-
opmental process (Bates and MacWhinney 1987; Bates et al. 1998; Elman
1999).
Captured by the term interaction engine (Levinson 2006), a second set
comprises social cognitive skills that include the basic motivation that
drives individuals to seek cooperative interaction with others, the abil-
ity to track actions and to infer the motivations, stances and intentions
behind them, and the ability to direct the attention of others and
project how one’s actions are understood by another. It also includes
the capacity for creating and interpreting communicative actions in
multiple modes, i.e., gesture, gaze, facial and other verbal channels,
simultaneously.

A third set of capacities involves what Tomasello (2003) refers to as

pattern-finding skills. These include the abilities to select and attend to
particular meaning-making components and to form perceptual and con-
ceptual categories of similar objects and events. Also included are skills to
form schemas based on recurrent patterns of action, to perform statisti-
cally based analyses on various kinds of patterns and sequences, and to
create mappings or analogies across two or more units based on functional
similarities.
Key aspects of our language experiences that give shape to language
knowledge are the distribution and frequency with which specific linguis-
tic resources are encountered in them. The more routine our everyday
activities are, and the more frequent, predictable and stable the use of
particular constructions is in them, the more entrenched the construc-
tions become as cognitive representations of our experiences. The more
entrenched they are, the more easily they are retrieved (Dabrowska 2008;
Rowland, Pine, Lieven and Theakston 2003; Tomasello 2003, 2006).
Also key to the development of language knowledge are cues used by
other participants that make salient and call attention to particular com-
ponents and their form–function relationships (Dabrowska 2012; Ellis and
Larsen-Freeman 2006; Tomasello 2003, 2006). These include explicit verbal
signals such as direct instructions, and less explicit verbal and non-verbal
cues such as repetition and tone and pitch changes, gaze, gesture and body
positionings. In classrooms, for example, such work is typically accom-
plished by both explicit and implicit instructional actions that direct
learners to attend to relevant resources and their form–meaning connec-
tions, and to make connections between them and their contexts of use.
The work of Joan Bybee (2006, 2010; Bybee and Hopper 2001b) has been
instrumental in revealing how frequency of use affects language knowl-
edge. According to Bybee, two types of frequency shape the construction of
one’s language knowledge. Type frequency has to do with the number of
distinct lexical items that can be substituted in a given slot in a pattern or
expression, while token frequency has to do with how often a particular
unit appears in the input.
Type frequency determines the productivity, generalizability and sche-
maticity of patterns (Ellis 2008). The frequency of a pattern that calls for
different items in particular slots strengthens the pattern, making it more
accessible for use with new items. An example of a highly frequent pattern
in English is the regular past tense ending, V+ ed. This is used to construct
the past tense of thousands of different verbs: e.g. help, helped; walk, walked;
call, called. The greater number of lexical items that occur in a certain
position in a pattern, the more likely a general category of items will be
created and the less likely the pattern will become associated with any
particular lexical item. Moreover, the more items a category covers, the
more general its defining features are and the more likely the category will
extend to new items (Bybee 2010; Ellis and Larsen-Freeman 2006).

Tokens are individual occurrences of units in concrete contexts. They

can include sounds, syllables, words, phrases and even entire sentences.
High token frequency promotes the entrenchment of particular units and
arrangement of units. Bybee (2006) refers to this as the conserving effect. As
an example, in English the trend is to regularize past tense endings of
irregular verb forms that occur infrequently. For example, clinged is more
often used as the past tense of cling rather than clung, which is considered to
be the grammatically prescribed form; creeped is often used as the past
tense of creep instead of crept, and weeped is often used as the past tense of
weep instead of wept. However, irregular past tense forms of verbs that are
used frequently (e.g. buy, bought; keep, kept; go, went; eat, ate) become stored
in one’s cognitive repertoire and are more easily remembered and
accessed.
This is also the case for particular co-occurring units. The higher their
frequency of use, the more likely they will be remembered and accessed as
whole units. Boyland’s (2001) study on the use of the English construction
X and I is a good example of this. In her corpus of public discussion groups
in the internet-based “usenet newsgroups” she found that the pronoun you
was the most frequently occurring pronoun with I in the construction X
and I. Because of its frequent use, she argues, speakers process the phrase
as one unit. Thus, they are more apt to use the unit you and I after preposi-
tions such as between and from rather than the grammatically prescribed
object pronoun me, as in between you and me. The higher the frequency of
tokens is, the stronger their cognitive representations are and the more
resistant they are to change (Diessel 2007). An extreme case of the conser-
ving effect is the autonomy effect. This refers to the learning of frequently
used strings of constructions by rote and storing them as whole units
without associating them with the units that comprise them. Examples
in English include the storing of phrases such as dog eat dog world as doggy
doggy world, and for all intents and purposes as for all intensive purposes.
Another effect of token frequency is the reducing effect. Highly used
phrases become automatized and tend to reduce phonetically. Examples
in English include gimme for give me, gonna for going to, hafta for have to, and I
dunno for I don’t know (Bybee 2006; Bybee and Scheibman 1999). According
to Bybee (2006), this happens because the articulatory representation of
words and sequences of words is made up of neuromotor routines. When
sequences of these routines are repeated, their execution becomes more
fluent and a new routine is established whereby a group of constructions is
articulated as a single unit.
Our language knowledge then consists of constructions that are
grounded in the expressions that we encounter regularly in our everyday
interactions. The more frequent the occurrence of different linguistic
types and tokens and the more our attention is drawn to their form–
meaning pairings, the more entrenched they become. Conversely, the
less frequent the occurrence of certain constructions and the less salient

they are, the weaker are their representations in our language knowledge
(Ibbotson 2013). While much of the research on the linkages between
language use, cognition and language development has focused on the
development of monolinguals, research on bilingual children shows a
similar role for frequency in the development of bilingual language knowl-
edge (Dabrowska 2008; De Houwer 2011; Hoff et al. 2012).1
There is no doubt that this body of evidence linking language usage to
language knowledge is compelling. It is, however, incomplete in explain-
ing fully the variability and diversity of bilingual and multilingual devel-
opment in that it does not account for the contextual, social and cultural
conditions that affect the types of linguistic experiences to which one is
exposed, and the various roles and role relationships one can take or be
assigned in these experiences, which in turn affect the frequency of
appearance and saliency of language patterns in the input. For this, we
need to look at the substantial body of linguistic ethnographic research on
language use.
9.2.2 Linguistic ethnographic account2

The foundation for much of this work can be traced back to the work of
noted linguistic anthropologist Dell Hymes (1972, 1974, 1992). A contem-
porary of Chomsky in the 1960s, Hymes reacted against the view of lan-
guage proposed by Chomsky as a fixed, universal property of the human
mind. He argued that it is too restrictive in that it did not account for the
social knowledge individuals rely on to produce and interpret utterances
appropriate to the particular contexts in which they occur. It is this social
knowledge, Hymes argued, that shapes and gives meaning to linguistic
forms. To be adequate, a theory of language must account for how “social
function gives form to the ways in which linguistic features are encoun-
tered in actual life” (Hymes 1974, p. 196).
Hymes thus called for a socially constituted linguistics, which takes
social function to be the source from which linguistic features are formed.3
A key concept of Hymes’ socially constituted linguistics is communicative
competence, defined as the knowledge of and ability to use language, and
thus, by definition, variable both within individuals across contexts and
within contexts across individuals. Hymes proposed this as an alternative
to the concept of linguistic competence, which was considered to exist
independently from any performance factors or non-linguistic cognitive
constraints.
To study the communicative competence of speech communities,
Hymes developed a research approach called the ethnography of commu-
nication. This was conceived as both a conceptual framework and a
method for conducting field studies of a community’s repertoire of com-
municative resources by which the community members engage in and
make sense of their communicative activities, and the larger social,

cultural, political and other institutional forces giving shape and meaning
to the resources. The analytical unit was conceived in terms of the com-
municative event. “In terms of language proper, [this] means that the
linguistic code is displaced by the speech act as the focus of attention”
(Hymes 1964, p. 13). Such studies for Hymes were to be “concrete, yet
comparative, cumulative, yet critical” (Hymes 1986, p. 63) with the overall
goal being not to seek “the replication of uniformity, but the organization of
diversity” (Hornberger 2009, p. 350). He noted, “the same behaviors, the
same verbal conduct, may have different implications for different actors . . .
What is a meaningful choice for one person may be the only way another
has of doing anything of the kind at all” (Hymes 1986, p. 87).
Although Hymes’ own work was not concerned directly with develop-
ment, two linguistic anthropologists, Elinor Ochs and Bambi Schieffelin,
drew on his theory of language and other research from the field of
developmental pragmatics (e.g. Ervin-Tripp and Mitchell-Kernan 1977) to
develop an approach to the study of language development they called
language socialization (Ochs and Schieffelin 1984, 1995, 2008; Ochs 1988,
2000). A key premise of this approach holds that language development is
embedded in and constitutive of the process of becoming a member of
one’s social groups. In their specific ways of using language in their com-
municative activities, expert members socialize children and other
novices into “a lifeworld saturated with social and cultural forces, predi-
lections, symbols, ideologies, and practices” (Ochs and Schieffelin 2008,
p. 5). In this process, the novices’ attention is directed to those aspects of
objects and events that are regularly encoded in the constructions of that
language.
As communities differ in the communicative goals they establish in
their activities, so do the communicative environments and the linguistic
resources used to socialize children and novices. The variability in
resources and opportunities for use leads to the development of individual
language knowledge characterized by variable “patchworks of function-
ally distributed communicative resources” (Blommaert and Backus 2011,
p. 9), which, in turn, lead to the formation of culture-specific habits of
thinking and construals of experiences (Slobin 1996, 2003). Ochs and
Schieffelin (2008) note that the process of socialization is not limited to
childhood but is a lifespan experience, characterizing “our human inter-
actions throughout adulthood as we become socialized into novel activ-
ities, identities and objects relevant to work, family, recreation, civic,
religious, and other environments in increasingly globalized commu-
nities” (p. 11).
Research by Bowerman and Choi (2001) provides a clear example of the
variability in language knowledge engendered by differences in socializa-
tion practices. Standard English uses the prepositions in and on to indicate
spatial relations of containment and support. In contrast, Korean distin-
guishes between loose fit and tight fit, with the verb kkita describing the

tight fitting together of two objects and the verb nehta indicating a loose
fit. In their examinations of spontaneous speech of English- and Korean-
speaking children between the ages of one and three, Bowerman and Choi
found that each group’s encoding of spatial configurations reflected the
major distinctions and grouping principles of their respective languages.
They argued that via their culture-specific processes of socialization, chil-
dren’s attention is directed to those aspects of objects and events that are
regularly encoded in the constructions of their specific languages.
Hymes’ theoretical insights and research methodologies paved the way
for decades of research concerned with uncovering the communicative
competence embodied in a vast range of language and literacy practices
found in social groups and communities around the world.4 A great deal of
this work has focused on the institution of schooling, addressing whether
and how in the varied communicative events comprising schools and their
surrounding communities, linguistic diversity is constructed as a problem
or resource, and the consequences in terms of learners’ language experi-
ences. An example of the varying outcomes in terms of language knowl-
edge engendered by socialization into different literacy practices across
home and school is Heath’s (1983) classic ethnographic investigation of
the literacy practices of three communities, two rural and one urban, all
located in the Carolina Piedmont region of the United States. Among other
findings, she found that the literacy practices between the two rural
communities into which children were socialized differed from each
other in fairly significant ways. For example, children in each community
were socialized into different understandings of the activity of “storytell-
ing” and into using different linguistic resources for accomplishing the
activity. Children in one community were encouraged to exaggerate and to
fantasize when telling their stories, whereas children from the other
community were expected to stick to the facts, to provide details where
necessary, but to never stray from what the adults considered to be “the
truth.” Moreover, it was found that the home literacy practices of both
communities differed from those found in the urban community, which
reflected those found in their schools. These differences across commu-
nities resulted in different literacy learning outcomes in school. Children
from the rural communities whose literacy practices into which they had
been socialized differed from their schools’ practices had more difficulty
succeeding academically than did their urban counterparts, whose home
socialization practices more closely resembled the literacy practices of
school.5
Studies exploring both the processes and products of language socializa-
tion have revealed the great variability of linguistic experiences into which
members of social groups are socialized. More specifically, the findings
provide rich empirical details on how the frequency and salience of lin-
guistic items that are used to socialize children and other novices, such as
L2 learners, in their linguistic experiences depend upon such social and

cultural factors as situational roles and rights of access, expectations about

participation, individual motivations, and so on. They are also subject to
larger social, historical and political forces, and these factors lead to varied
trajectories of individual language development, even within the same
linguistic experiences.6
Talmy’s (2008) ethnographic study of ESL classes illustrates how institu-
tional expectations about ESL learners can lead to differentially organized
classroom-based socialization practices. In his investigation of three first-
year high school ESL classes in school located in the US state in Oceania, he
revealed that although the student population in these classes included
students local to the area, who thus had experience with US schooling
contexts and practices, classroom practices oriented to the learners as
recently arrived cultural and linguistic “others” (p. 626). Literacy practices,
for example, included L1 English juvenile fiction books such as James and
the Giant Peach (Dahl 1961), which were below grade-level and had only
indirect relevance to high school academics or L2 learning. This mismatch
between the ESL student identity embodied in the official institutional
curriculum and the local ESL students’ self-identifications led to the devel-
opment of a repertoire of classroom practices that marked a resistance to
and a lack of investment in the official curriculum. This in turn led to a
substantial decline in academic performance over the school year.
In contrast to the study showing how socialization practices arising
from competing ideologies can lead to disenfranchisement and lack of
advancement for some is the study by Creese et al. (2006). Their investiga-
tion of two Gujarati complementary schools in the United Kingdom
reveals how in maintaining particular heritages, community traditions
and languages in their institutional practices, the schools promoted a
more expansive space for the construction of student identities, affording
some students the opportunity to position themselves as multicultural and
multilingual, rather than bound to one language or culture.
Continuing the theoretical commitment to advancing understandings
of diversity in competence, contemporary work is motivated by a desire to
understand the multilingual and multimodal communicative practices
resulting from global migration and the urbanization of communities.
The social landscapes of communities are now super-diverse (Vertovec
2007, 2010), characterized by social groupings comprising diverse nation-
alities, ethnicities, religious traditions, languages and cultural values and
practices. Makoni et al. (2007) provide a compelling study of the effects of
rapid urbanization on the language practices of multilingual commu-
nities. Their context is Harare, Zimbabwe, which, like other urban loca-
tions in Africa, is characterized by multilingual interactions in which
speakers understand a range of local languages and are willing to accom-
modate the speakers of the different languages. Detailed linguistic analy-
ses of these practices reveal that the linguistic resources that speakers
draw on in these interactions are best characterized as “urban

vernaculars” (Makoni et al. 2007, p. 34), variable collections made up of

constructions that are drawn from a number of different languages.
Individuals share these vernaculars to the extent that they “share similar
social and economic experiences” (p. 42), with differences depending on
individuals’ social and urban networks.
Additional research shows the significant role that ever-expanding digi-
tal communication and information technologies such as the computers,
video games, smart phones and the internet play, not only in making
possible interactions with geographically distant communities. They
have also expanded available semiotic resources to include the graphic,
pictorial, audio, physical and spatial (Blommaert 2010; Coupland 2010).
These ever-expanding opportunities for creating, exchanging and inter-
preting information and building social networks have increased the
scope and constitution of communication practices between individuals
and within and across social groups and communities worldwide.
A case study by Lam (2009) illustrates the types of multilingual and
multimodal practices that can arise from transnational migration and
the new forms of communication. Her study reveals how a young
Chinese woman who had immigrated to the United States was able to
assemble a variety of multilingual resources from her transnational net-
works to develop affiliations with different communities across geogra-
phical borders. As one example of the resources she assembled, in order to
construct a English-speaking voice for herself and to develop an affiliation
with Asian-Americans who identified with an urban youth culture in
various locations across North America, the young woman created a
hybrid variety of English that incorporated the standard orthography of
written English, the non-standard orthography of African American
Vernacular English and hip-hop, and the conventions of instant messaging
in her interactions with on-line friends she had become affiliated with
through the playing of an on-line game.
Alongside mounting evidence on the ever-increasing diversity of multi-
lingual and multimodal resources comprising contemporary practices,
new concepts have been proposed to more aptly explain the processes by
which individuals develop and use them. The term repertoire appears to be
gaining traction in the literature. Critics note that the term language
carries an ideology that implies homogeneity and permanence
(Blommaert and Rampton 2011; Makoni et al. 2007). In contrast, reper-
toire, defined as dynamic and malleable “constellations of semiotic
resources” (Hall et al. 2006, p. 232) that individuals draw on and develop
from their specific contexts of use, more aptly captures understandings of
individual competence as emergent and provisional.7
Blommaert and Backus (2011) expand on the term, arguing that the
origins of repertoires are biographical, and thus historical, a product of
the diversity of learning environments inhabited by individuals over their
lifespans. Development happens “explosively in some phases of life and

gradually in some others” (p. 8), with resources learned in a “wide variety of
trajectories, tactics and technologies, ranging from fully formal language
learning to entirely informal ‘encounters’ with language” (p. 1), with the
varied conditions leading to fuller and more enduring competence in some
contexts and more temporary, fleeting competence in others.
Another term gaining traction in the field is translanguaging, to refer the
processes by which individuals make use of their repertoires. Garcia (2009)
defines it as the “multiple discursive practices in which bilinguals engage
in order to make sense of their bilingual worlds”(p. 45, original emphasis).
Related terms include crossing (Rampton 2005), codemeshing (Canagarajah
2011) and metrolingualism (Otsuji and Pennycook 2010). Li (2011) proposes
an alternative definition that embraces a more spontaneous, creative and
critical view. For Li Wei translanguaging is:
both going between different linguistic structures and systems and going
beyond them. It includes the full range of linguistic performances of
multilingual language users for purposes that transcend the combination
of structures, the alternation between systems, the transmission of infor-
mation and the representation of values, identities and relationships. The
act of translanguaging then is transformative in nature; it creates a social
space for the multilingual language user by bringing together different
dimensions of their personal history, experience and environment, their
attitude, belief and ideology, their cognitive and physical capacity into one
coordinated and meaningful performance, and making it into a lived
experience. (Li Wei 2011, p. 1223)
What joins the various terms is their focus on “how different communica-
tive resources are employed to create meaning and what such a hetero-
glossic language practice means to speakers” (Busch 2012, p. 505).
Additional terms gaining purchase include communities of practice, knowledge
communities and social networks to describe groupings of individuals who
come together for shared purposes organized around, for example, social
or professional goals.
In sum, the theoretical and methodological insights on the socially con-
stituted nature of competence, proposed over half a century ago by Hymes
and his colleagues, has yielded compelling empirical evidence on the diver-
sity of resources individuals draw on to participate in and make sense of
their world, the contextual, social and cultural conditions that influence
their use, and the consequences that participation in their experiences have
for individual action and development. Together with findings uncovering
the role played by cognition in the development of these resources, they
give substantial empirical weight to an account of multi-competence as “a
minimally sorted and organized set of memories of what people have heard
and repeated over a lifetime of language use, a set of forms, patterns, and
practices that have arisen to serve the most recurrent functions that speak-
ers find need to fulfil” (Ford et al. 2003, p. 122).

9.3 Moving forward
Two broad implications for advancing research on multi-competence can

be drawn from the usage-based account presented in this chapter. The first
is specific to the linguistic ethnographic research traditions. In addition to
yielding new insights on language use and development, the proliferation
of studies over recent years has witnessed an increase in the number of
new or blended terms used to describe various phenomena. As noted
earlier, included are terms such as translanguaging, metrolingualism
and crossing. An advantage of these terms is that they come without
histories of meanings and so can be imbued with a signification that
represents new ways of seeing and understanding experiences. Other
terms such as communities of practice, repertoire, style and register are
linked to multiple bodies of knowledge which thus come with historical
attachments to different disciplines and so with meanings that are fairly
flexible and malleable. Other terms such as language and grammar have
long attachments to one discipline, making their histories invisible and
their meanings almost impossible to change.
Why does this matter? A primary reason for undertaking research on
language use in society is to develop more comprehensive ways of seeing
the formative links between our social worlds and individual multi-com-
petence. This entails developing shared knowledge practices that can
focus researchers’ efforts into the systematic pursuit of knowledge
advancement. Concepts are the basic tools for thinking and inquiry.
They organize and synthesize research findings so as to enhance under-
standings and awareness of the phenomena. Essential to the development
of shared research practices then is the development of a coherent set of
concepts. As exciting as the proliferation of new terms can seem, it can
also be a detriment to the advancement of knowledge. Using the same
terms to refer to different experiences in different contexts and different
terms to refer to seemingly similar experiences can lead to conceptual
confusion. We saw this earlier in the different meanings ascribed to
translanguaging, and the use of translanguaging and code-meshing to
refer to similar phenomena.
Using terms from other disciplines can be challenging as well as. As
noted earlier, the concept of repertoire has become commonplace in
current studies, and Busch (2012) provides a useful summary of its history
of meaning in other research arenas as a background to her use of the
term in her study. However, it is not clear in her use how the term relates
to the meaning of language and how it relates to the concept of multi-
competence, if it does at all.
The term multi-competence is another example of the challenges that
arise when using concepts with past histories of meaning to refer to new
understandings. As noted earlier, Hymes’ initial proposal of the term

communicative competence was meant to capture the socially contingent

and dynamic nature of language. As Franceschini (2011) notes, “commu-
nicative competence can easily comprehend different and typologically
distant varieties. In this way, the term is open to include multilingual
competencies” (p. 347). However, despite the capacity of communicative
competence to capture the diversity of linguistic resource, she argues for
the term multi-competence, as it more usefully “captures multilingualism
as a culturally, historically, and dynamically determined means of com-
munication” (p. 350).
This may be so. It is certainly the case that the complexity of the concept
of communicative competence as articulated by Hymes has never been
fully appreciated, despite the richness of findings yielded by the work of
Hymes and others. As noted by Blommaert (2009), the version of commu-
nicative competence commonly found in the field of applied linguistics
has accumulated meanings that bear “only the vaguest and most distant
traces of its Hymesian origins” (p. 261), thus rendering the term unsuitable
for use in contemporary work.8 Replacing the term communicative with the
prefix multi- is a more explicit reminder of the heterogeneity of language
knowledge, and thus can help to work against any lingering monolingual
bias in research endeavours (Ortega 2013).
However, multi-competence comes with its own history of meaning (see
Hall et al. 2006), which needs to be unpacked and understood in order to
use it to address current understandings in ways that avoid conceptual
confusion. Franceschini (2011) is a case in point. In her argument for
integrating the notion of multilingualism with multi-competence, she
provides two competing definitions of the latter term. According to
Franceschini, multi-competence is “the knowledge of more than one lan-
guage in the mind” (p. 351), and a multi-competent person is one “with
knowledge of an extended and integrated linguistic repertoire who is able
to use the appropriate linguistic variety for the appropriate occasion”
(ibid.). The latter definition is more attuned to current empirical findings,
while the former definition runs counter to current understandings (see
Hall et al. 2006). This is an instructive example of how, despite one’s stated
theoretical leanings, it can be difficult to let go of old habits of “thinking
for speaking” (see Slobin 2003). If multi-competence is to remain a viable
concept, those using it must be mindful to relinquish its past meanings
and more fully embrace the usage-based perspective.
This points to the even greater challenge of using concepts like language
and grammar, whose usage is so established and thus their meanings so
entrenched that changing them to mean something else may be impossi-
ble, despite our repeated contestations that they be changed. Indeed, as
Blommaert and Rampton (2011, p. 4) note, even with research agendas that
aim “to challenge nation-state monolingualism, languages are sometimes
still conceptualized as bounded [abstract] systems linked with bounded
communities.”

Motivated by a desire to contribute to the creation of new terms to

capture current realities, Blommaert (2012, p. 2) offers the term super-
vernaculars to describe “new forms of semiotic codes emerging in the
context of technology-driven globalization processes.” While it captures
the types of resource frequently encountered in technologically mediated
communicative practices, it does not account for those encountered in
other, equally diverse encounters. Whether this term suffices as a replace-
ment for the term “language” is an open question (see Orman 2012 for a
critique of the term). At the least, it opens the discussion on the need for
terms that offer more insightful understandings of the varied communi-
cative resources we encounter in our social worlds.
Alongside a coherent set of concepts, we need a conceptual framework
that organizes the concepts such that their similarities and distinctions are
clear, thereby allowing for analyses that are more interpretive and expla-
natory than merely descriptive. As a case in point, a good deal of recent
research has focused on describing multilinguals’ creativity in making use
of their linguistic resources (e.g. Li Wei 2011; Otsuji and Pennycook 2010).
Rich descriptions of these moments can help us see something we might
not have seen otherwise. However, without some principled ways of dis-
tinguishing creativity from conventionality and, concomitantly, differ-
ence from deficiency, it is difficult to know the source of the novelty.
This may not be significant in studies of informal contexts of use, where
individual uses are usually socially inconsequential. But it matters greatly
in formal learning and professional contexts of use, where differences in
learning opportunities and access to their affordances within the oppor-
tunities can mean great success for some, and failure for others. A sound
conceptual framework organized around a coherent set of concepts can
help us to separate out the material conditions that facilitate multilingual
development from those that constrain it.
A second broad challenge has to do with research design and imple-
mentation. While the issues raised are addressed primarily to the
linguistic ethnographic traditions, I note where they also concern the
cognitive approaches. In terms of design, much of the research moti-
vated by Hymes’ work has been more concerned with identifying the
communicative competence shared by members of groups and com-
munities and less with identifying the range of individual diversity
within groups. While we have learned a great deal about the commu-
nicative practices that shape and are shaped by social groups, we know
less about individual histories. The more recent focus on individual
repertoires as a unit of analysis will help expand the direction of
research attention to include the study of individuals and their his-
tories, assuming, of course, that the conceptual underpinnings of the
term can be clarified. Studies of individual repertoires will enable us to
document in greater detail their trajectories, including the opportu-
nities and constraints they face in terms of access to learning

environments and their resources and their movements within and

across contexts.
Although not directly focused on documenting the development of reper-
toires, current work examining multilingual identities and subjectivities
(e.g. Block 2007; Kramsch 2009) show the value of using case studies,
biographies and interviews to explore aspects of individual repertoire devel-
opment and use. Busch’s study (2012) is more directly concerned with
linguistic repertoires. She uses language portraits as a method to explore
how speakers depict their linguistic repertoires through visual and narra-
tive descriptions. All of these methods are useful in capturing individual
insights on their repertoires but they are only supplemental. To be able to
track individual trajectories of repertoires as they unfold within and across
an individual’s experiences, identify changes and link them to specific
contextual conditions arising as the trajectories unfold we need more
micro-analytic, ontogenetic and micro-genetic studies of human action.
The ever-expanding interest in understanding multilingual practices
and the individual repertoires that give shape to and arise from these
practices calls for greater attention to methods of data collection and
analysis, since the methods we choose shape the kinds of data we gather
and ultimately what we find. Because the ethnographic tradition begun by
Hymes requires that data be taken from naturally occurring events, data
collection methods often involve video- and audio-recordings. Prior to
recording any event, decisions must be made about the types of recording
equipment to use and their locations relative to the individuals and the
event being recorded. Additional decisions may be made during the record-
ing as the event unfolds, to take into account its emergent nature. These
decisions have been shown to play a significant role in shaping the event as
it is being recorded and thus are deeply implicated in the processes of
analysis (Mondada 2007).
Once recorded, the event is then transcribed and annotated. While on
the surface the process of representing events graphically may seem
unproblematic, it is a complex, theoretical task (Ochs 1979). In the process
of transcribing, we select actions from a much larger repertoire to repre-
sent in another form. In turn, the specific aspects we select shape what we
determine to be relevant in the analysis. For example, choosing to use
conventional orthography to represent oral discourse in English renders
invisible vocal qualities of the discourse such as pitch ranges and intona-
tional contours and temporal aspects such as pauses and gaps, all of
which have been shown to be essential cues in assigning meaning to verbal
cues. Moreover, our actions are never limited to just the verbal mode.
Gestures, body positionings, gaze and other facial cues shape the way in
which our actions are produced and understood. Not including them in the
transcript further constrains what our analysis can yield.9
These transcription issues are complicated enough when dealing with
events where a single conventional code is used, e.g. standard American

English; participants have shared repertoires and knowledge of the event

and transcription conventions for capturing the multimodal layers of
actions are well established. Researchers working with multilingual
events face unique challenges in producing faithful transcripts, even
more so if the events bring together individuals with different back-
grounds and repertoires. Decisions as to what and how to record and
transcribe will be highly consequential to what we learn about multi-
competence, and will thus require researchers to be highly skilled in
representing their data such that the claims made are ecologically valid,
i.e. fair and credible representations of the social worlds that are being
examined.
While multilingual practices have become a central focus for the lin-
guistic ethnographic traditions, this has not been the case for the cogni-
tive-oriented approach. Indeed, a great deal of what we know about the
linkages between cognitive processes, competence and language knowl-
edge is based on monolingual populations who live in fairly stable social
environments. For their findings to be relevant to multilingual popula-
tions, the data they draw on need to be more diverse. Linguistically diverse
communities such as the one researched by Makoni et al. (2007) need to
become the norm rather than the exception. Incorporating data from
multilingual populations into their research agenda will require multi-
lingual researchers who know the repertoires and can render faithful
transcriptions of oral, multimodal and textual practices in ways that
allow for the identification of tokens, formulaic expressions, patterns
and so on.
9.4 Conclusions
The usage-based account of multi-competence presented here brings

together findings from cognitive-oriented research focused on the lin-
kages between cognitive processes, linguistic input and language knowl-
edge, and the ethnographically oriented research focused on linkages
between language knowledge, linguistic input and the larger cultural,
institutional and historical conditions giving shape to the linguistic
input. Neither strand should be considered more essential to explanations
of multi-competence. Indeed, as Ibbotson (2013, p. 13) notes, when find-
ings from multiple independent methods are in agreement they can “con-
verge to strong conclusions.” Such a state is a good example of
“reductionism in the better sense of the word: where each domain of
understanding is grounded in the next but with no domain being dispen-
sable and where the emphasis is on unifying, integrating or synthesizing
aspects of knowledge” (ibid.). The theoretical alliance of the two
approaches is a beginning. As we move forward, a challenge for furthering
understandings of multi-competence will be to integrate more fully the

empirical findings from both. This will require a fairly strong commitment
from both camps to find ways, conceptually, methodologically and other-
wise, to come together in “the dogged work of making sense of real
communities and real lives” (Hymes 1974, p. 7).
Notes
1. For a far more comprehensive account of the cognitive-oriented

research that underpins a usage-based approach, taking into account
research on both monolinguals and L2 learners, readers are directed to
several summary accounts by Nick Ellis such as Ellis (2008), Ellis (2012)
and Ellis et al. (2013).
2. This term comes from Rampton (2007), who provides a useful summary
of various research traditions concerned with the analysis of language in
society that have developed over the last five to fifteen years in the UK,
and links it to the traditions influenced by Hymes’ work in the US over
the last 40 or so years. A hallmark of these traditions, he notes, is their
shared commitment to “putting linguistics and ethnography together to
try to understand the social processes that we are involved in” (p. 599). He
uses the term “linguistic ethnography” to characterize this on-going
interdisciplinary research activity. It seems appropriate to use it here.
3. It should be noted that as Hymes, an American linguistic anthropolo-
gist, was developing his theory of language, Michael Halliday, a British
linguist mostly working in Australia, was developing a similar theory of
language as a sociocultural resource constituted by “an open-ended set
of options in behaviour that are available to the individual in his
existence as social man” (Halliday 1973, p. 49). However, they developed
different approaches to the study of language, with Hymes taking an
ethnographic approach and Halliday a social semiotic approach.
4. See Bauman and Sherzer (1974) for examples of studies; see Phillipsen
and Carbaugh (1986) for a bibliography of over 150 studies.
5. As Hornberger (2009) notes, linguistic inequality, especially as it relates
to language education, was a lasting theme of Hymes’ work. In addition
to the groundbreaking study by Heath (1983) mentioned earlier, are
those by Phillips (1983), Saville-Troike (1987) and Cazden (1988), to
name a very few.
6. In her review of language socialization, Ervin-Tripp (2009) notes two
strands of research deriving from Hymes’ theory, and both, she argues,
need further elaboration. The first strand emphasizes the role of lan-
guage in the acquisition of culture in different cultural settings but
needs more systematic comparison across sites. The second focuses on
documenting changes in linguistic resources brought about by sociali-
zation, but needs more explanatory attention to the organization of the
linguistic means in the service of contextual outcomes.

7. This is not a new term. In fact it has a long history of use in linguistic
anthropology. See Blommaert and Backus (2011) and Busch (2012) for
reviews of its history and arguments for its relevance to current under-
standings of language knowledge.
8. Canale and Swain (1980) brought the term to the field of SLA for the
purposes of redesigning L2 pedagogy and assessment, and in those
arenas it has provided a useful heuristic for designing curriculum and
assessment materials. However, it has had little to no influence on
research into L2 learning.
9. See Bucholtz (2007) for a compelling example of how different inter-
pretations of actions can arise from variation in transcriptions.
References
Bates, E. and MacWhinney B. 1987. ‘Competition, variation, and language

learning’, in B. MacWhinney (ed.), Mechanisms of Language Acquisition,
pp. 157–194. Hillsdale, NJ: Lawrence Erlbaum.
Bates, E., Elman, J., Johnson, M., Karmiloff-Smith, A., Parisi, D. and Plunkett,
K. 1998. ‘Innateness and emergentism’, in W. Bechtel and G. Graham
(eds), A Companion to Cognitive Science, pp. 590–601. Oxford: Basil Blackwell.
Bauman, R. and Sherzer, J. (eds) 1974. Explorations in the Ethnography of
Speaking. Cambridge University Press.
Blackledge, A. 2005. Discourse and Power in a Multilingual World. Amsterdam:
John Benjamins.
Block, D. 2007. Multilingual Identities in a Global City: London Stories. Palgrave
Macmillan.
Blommaert, J. M. 2009. ‘Ethnography and democracy: Hymes’s political
theory of language’, Text and Talk 29, 257–276.
Blommaert, J. M. 2010. The Sociolinguistics of Globalization. Cambridge
University Press.
Blommaert, J. M. 2012. ‘Supervernaculars and their dialects’, Dutch Journal
of Applied Linguistics 1, 1–13.
Blommaert, J. M. and Backus, A. 2011. ‘Repertoires revisited: “Knowing
language” in superdiversity’, Working Papers in Urban Language and
Literacies 67, downloaded July 23, 2012. www.kcl.ac.uk/sspp/depart
ments/education/research/ldc/publications/workingpapers/67.pdf
Blommaert, J. M. and Rampton, B. 2011. ‘Language and superdiversity’,
Diversities 13, 3–21.
Bowerman, M. and Choi, S. 2001. ‘Shaping meanings for language: uni-
versal and language-specific in the acquisition of spatial semantic cate-
gories’, in M. Bowerman and S. C. Levinson (eds), Language Acquisition and
Conceptual Development, pp. 475–511. Cambridge University Press.
Boyd, J. and Goldberg, A. 2009. ‘Input effects within a constructionist
framework’, Modern Language Journal 93, 418–429.

Boyland, J. 2001. ‘Hypercorrect pronoun case in English? Cognitive pro-

cesses that account for pronoun usage’, in J. Bybee and P. Hopper (eds),
pp. 383–404.
Bucholtz, M. 2007. ‘Variation in transcription’, Discourse Studies 9, 784–808.
Busch, B. 2012. ‘The linguistic repertoire revisited’, Applied Linguistics 33,
503–523.
Bybee, J. 2006. ‘From usage to grammar: the mind’s response to repetition’,
Language 82, 711–733.
Bybee, J. 2010. Language, Usage and Cognition. Cambridge University Press.
Bybee, J. and Hopper, P. (eds) 2001a. Frequency and the Emergence of Linguistic
Structure. Amsterdam: John Benjamins.
Bybee, J. and Hopper, P. 2001b. ‘Introduction to frequency and the emer-
gence of linguistic structure’, in J. Bybee and P. Hopper (eds), pp. 1–26.
Bybee, J. and Scheibman, J. 1999. ‘The effect of usage on degrees of
constituency: the reduction of don’t in English’, Linguistics 37 (4),
575–596.
Canagarajah, A. S. 2011. ‘Codemeshing in academic writing: identifying
teachable strategies of translanguaging’, Modern Language Journal 95,
401–417.
Canale, M. and Swain, M. 1980. ‘Theoretical bases of communicative
approaches to second language teaching and testing’, Applied Linguistics
1, 1–47.
Cazden, C. 1988. Classroom Discourse: The Language of Teaching and Learning.
Portsmouth, NH: Heinemann.
Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.
Collins, J. 2014. ‘Literacy practices, linguistic anthropology and social
inequality’, Working Papers in Urban Language and Literacies 143, down-
loaded November 15, 2014. https://www.academia.edu/9277086/
WP143_Collins_2014._Literacy_practices_linguistic_anthropology_and_
social_inequality.
557–591.
Coupland, N. (ed.) 2010. Handbook of Language and Globalization. Malden, MA:
Wiley-Blackwell.
Creese, A., Bhatt, A., Bhojani, N. and Martin, P. 2006. ‘Multicultural, heri-
tage and learner identities in complementary schools’, Language and
Education 20, 23–43.
Da˛browska, E. 2008. ‘The effects of frequency and neighbourhood density
on adult native speakers’ productivity with Polish case inflections: an
empirical test of usage-based approaches to morphology’, Journal of
Memory and Language 58, 931–951.

Da˛browska, E. 2012. ‘Different speakers, different grammars: individual

differences in native language attainment’, Linguistic Approaches to
Dahl, R. 1961. James and the Giant Peach. New York: Knopf.
De Houwer, A. 2011. ‘Language input environments and language devel-
opment in bilingual acquisition’, Applied Linguistics Review 2, 221–240.
Diessel, H. 2007. ‘Frequency effects in language acquisition, language use,
and diachronic change’, New Ideas in Psychology 25, 108–127.
Ellis, N. 2006. ‘Cognitive perspectives on SLA: the Associative–Cognitive
CREED’, AILA Review 19, 100–121.
Ellis, N. 2008. ‘Usage-based and form-focused language acquisition: the
associative learning of constructions, learner-attention and the limited
L2 endstate’, in P. Robinson and N. Ellis (eds), Handbook of Cognitive
Linguistics and Second Language Acquisition, pp. 372–405. London: Routledge.
Ellis, N. 2012. ‘What can we count in language, and what counts in lan-
guage acquisition, cognition, and use?’, in S. T. Gries and D. S. Divjak
(eds), Frequency Effects in Language Learning and Processing, pp. 7–34. Berlin:
Mouton De Gruyter.
Ellis N. and Larsen-Freeman, D. (eds) 2006. Language Emergence: Implications
for Applied Linguistics, special issue of Applied Linguistics 27.
Ellis, N., O’Donnell, M. and Romer, U. 2013. ‘Usage-based language: inves-
tigating the latent structures that underpin acquisition’, Language
Learning 63, 25–51.
Elman, J. 1999. ‘Origins of language: a conspiracy theory’, in B.
MacWhinney (ed.), The Emergence of Language, pp. 1–28. Hillsdale, NJ:
Lawrence Erlbaum.
Ervin-Tripp, S. 2009. ‘Hymes on speech socialization’, Text and Talk 19,
245–256.
Ervin-Tripp, S. and Mitchell-Kernan, C. (eds) 1977. Child Discourse. Academic
Press, New York.
Ford, C., Fox, B. and Thompson, S. 2003. ‘Social interaction and grammar’,
in M. Tomasello (ed.), The New Psychology of Language, vol. 2, pp. 119–143.
Mahwah, NJ: Lawrence Erlbaum.
Garcia, O. 2009. Bilingual Education in the 21st Century: A Global Perspective,
Oxford: Blackwell.
Gumperz, J. J. 1982. Discourse Strategies. Cambridge University Press.
Hall, J. K., Cheng, A. and Carlson, M. T. 2006. ‘Reconceptualizing multi-
220–240.
Halliday, M. A. K. 1973. Explorations in the Functions of Language. London:
Edward Arnold.
Halliday. M. A. K. 1975. Learning How to Mean: Explorations in the Development of
Language. London: Edward Arnold.

Halliday. M. A. K. 1978. Language as a Social Semiotic: The Social Interpretation of

Language and Meaning. London: Edward Arnold.
Hanks, W. 1996. Language and Communicative Practices. Boulder, CO:
Westview Press.
Heath, S. B. 1983. Ways with Words: Language, Life, and Work in Communities
and in Classrooms. Cambridge University Press.
Hoff, E., Core, C., Place, S., Rumiche, R., Señor, M. and Parra M. 2012. ‘Dual
language exposure and early bilingual development’, Journal of Child
Language 39, 1–27.
Hornberger, N. 2009. ‘Hymes’s linguistics and ethnography in education’,
Text and Talk 29, 347–358.
Hymes, D. 1964. ‘Formal discussion’, The Acquisition of Language: Monographs
of the Society for Research in Child Development 29, 107–111.
Hymes, D. 1972. ‘On communicative competence’, in J. B. Pride and
J. Holmes (eds), Sociolinguistics, pp. 269–293. Harmondsworth: Penguin.
Hymes, D. 1974. Foundations in Sociolinguistics: An Ethnographic Approach.
Philadelphia: University of Pennsylvania Press.
Hymes, D. 1986. ‘Discourse: scope without depth’, International Journal of the
Sociology of Language 57, 49–89.
Hymes, D. 1992. ‘The concept of communicative competence revisited’, in
M. Putz (ed.), Thirty Years of Linguistic Evolution, pp. 31–58. Philadelphia:
John Benjamins.
Ibbotson, P. 2012. ‘A new kind of language’, The Psychologist 25 (2), 122–125.
Ibbotson, P. 2013. ‘The scope of usage-based theory’, Frontiers in Psychology 4,
1–15.
Kramsch, C. 2009. The Multilingual Subject. Oxford University Press.
Lam, W. S. E. 2009. ‘Multiliteracies on instant messaging in negotiating
local, translocal, and transnational affiliations: a case of an adolescent
immigrant’, Reading Research Quarterly 44, 77–397.
Langacker, R. 1987. Foundations of Cognitive Grammar, vol. 1. Stanford, CA:
Stanford University Press.
Lee, N., Mikesell, L., Joaquin, A., Mates, A. and Schumann J. 2009. The
Interactional Instinct: The Evolution and Acquisition of Language. Oxford
University Press.
Levinson, S. C. 2006. ‘On the human “interaction engine”’, in
N. J. Enfield and S. C. Levinson (eds), Roots of Human Sociality,
pp. 39–69. Oxford: Berg.
Li Wei 2011. ‘Moment analysis and translanguaging space: discursive con-
struction of identities by multilingual Chinese youth in Britain’, Journal
of Pragmatics 43, 1222–1235.
Liddicoat, A. J. and Crichton, J. 2008. ‘The monolingual framing of inter-
national education in Australia’, Sociolinguistic Studies 2, 367–384.
MacWhinney, B. 2008. ‘A unified model’, in P. Robinson and N. Ellis (eds),
Handbook of Cognitive Linguistics and Second Language Acquisition, pp. 341–371.

Makoni, S., Brutt-Griffler, J. and Mashiri, P. 2007. ‘The use of “indigenous”

and urban vernaculars in Zimbabwe’, Language in Society 36, 25–49.
Makoni, S. and Pennycook, A. (eds) 2007. Disinventing and Reconstituting
Languages. Clevedon, UK: Multilingual Matters.
Mondada, L. 2007. ‘Multimodal resources for turn-taking: pointing and the
emergence of possible next speakers’, Discourse Studies 9, 194–225.
Ochs, E. 1979. ‘Transcription as theory’, in E. Ochs and B. B. Schieffelin
(eds), Developmental Pragmatics, pp. 43–72. New York: Academic Press.
Ochs, E. 1988. Culture and Language Development: Language Acquisition and
Language Socialization in a Samoan Village. Cambridge University Press.
Ochs, E. 2000. ‘Socialization’, Journal of Linguistic Anthropology 9, 230–233.
Ochs, E. and Schieffelin, B. 1984. ‘Language acquisition and socialization:
three developmental stories’, in R. A. Shweder and R. A. LeVine (eds),
Culture Theory: Essays on Mind, Self, and Emotion, pp. 276–320. Cambridge
University Press.
Ochs E. and Schieffelin, B. 1995. ‘The impact of language socialization on
grammatical development’, in P. Fletcher and B. MacWhinney (eds), The
Handbook of Child Language, pp. 73–94. Oxford: Blackwell.
Ochs, E. and Schieffelin, B. 2008, ‘Language socialization: an historical
overview’, in P. A. Duff and N. Hornberger (eds), Encyclopedia of Language
and Education, second edition, pp. 3–15. Springer.
Orman, J. 2012. ‘Not so super: the ontology of ‘supervernaculars’, Language
and Communication 32 (4), 349–357.
Ortega, L. 2013. ‘Ways forward for a bi/multilingual turn in SLA’, in S. May
(ed.), The Multilingual Turn: Implications for SLA, TESOL and Bilingual
Education, pp. 32–52. New York: Routledge.
Otsuji, E. and Pennycook, A. 2010. ‘Metrolingualism: fixity, fluidity and
language in flux’, International Journal of Multilingualism 7, 240–254.
Philips, S. 1983. The Invisible Culture: Communication in Classroom and
Community in the Warm Springs Indian Reservation. White Plains, NY:
Longman.
Philipsen G. and Carbaugh, D. 1986. ‘A bibliography of fieldwork in the
ethnography of communication’, Language in Society 15, 387–398.
Rampton, B. 2005. Crossing: Language and Ethnicity among Adolescents, second
edition. Manchester, UK: St Jerome Press.
Rampton B. 2007. ‘Neo-Hymesian linguistic ethnography in the United
Kingdom’, Journal of Sociolinguistics 11, 584–607.
Rowland, C., Pine, J., Lieven E. and Theakston A. 2003. ‘Determinants of the
order of acquisition of wh-questions: re-evaluating the role of caregiver
speech’, Journal of Child Language 30, 609–635.
Saville-Troike, M. 1987. ‘Dilingual discourse: the negotiation of meaning
without a common code’, Linguistics 25, 81–109.
Slobin, D. 1996. ‘From “thought” and “language” to “thinking for
speaking”’, in J. J. Gumperz and S. C. Levinson (eds), Rethinking Linguistic
Relativity, pp. 70–96. Cambridge University Press.

Slobin, D. 2003. ‘Language and thought online: cognitive consequences of

linguistic relativity’, in D. Gentner and S. Goldin-Meadow (eds), Language
in Mind: Advances in the Study of Language and Thought, pp. 157–192.
Cambridge, MA: MIT Press.
Talmy, S. 2008. ‘The cultural productions of ESL student at Tradewinds
High: Contingency, multidirectionality, and identity in L2 socialization’,
Tomasello, M. 2003. Constructing a Language: A Usage-based Theory of Language
Acquisition. Harvard, MA: Harvard University Press.
Tomasello, M. 2006. ‘Acquiring linguistic constructions’, in D. Kuhn and R.
Siegler (eds), Handbook of Child Psychology, vol. 2, pp. 255–298. Oxford
University Press.
Vertovec S. 2007. ‘Super-diversity and its implications’, Ethnic and Racial
Studies 30, 1024–1054.
Vertovec, S. 2010. ‘Towards post-multiculturalism? Changing commu-
nities, contexts and conditions of diversity’, International Social Science
Journal 199, 83–95.
Vygotsky, L. S. 1978. Mind in Society. Cambridge, MA: Harvard University
Press.

10
Multi-competence
and syntax
Éva Berkes and Suzanne Flynn
10.1 Introduction
From antiquity to the present century we may easily find evidence for
overwhelming bi- and plurilingualism. ‘Language contact intruded
into virtually every aspect of ancient life: e.g., high literature, the law,
medicine, magic, religion, provincial administration, the army, and trade’
(Adams and Swain 2002, p. 1). Given the increased pace of globalization
and the boom in the use of social networking sites in the last ten years,
as seen in the Pew Center’s Report (2013), the tendency to grow up in a
plurilingual community and actively use multiple languages on a daily
basis is likely to increase. See similar arguments in Chomsky (2006),
Grosjean (2004), Auer and Wei (2007), Cook and Newson (2007) and
Aronin and Singleton (2012), among others. Bhatia and Ritchie, in their
recent Handbook of Bilingualism and Multilingualism, claim that ‘far from
being exceptional, as many laymen believe, plurilingualism . . . is currently
the rule throughout the world and will become increasingly so in the
future’ (Bhatia and Ritchie 2013). National borders may, in some cases,
still represent the borders of a linguistic community, but even these
communities are affected by migration of other linguistic communities
due to sociopolitical, economic or even climatic reasons. Increased infra-
structure, better communication facilities and the explosive growth of
media have not only opened up but practically erased barriers among
speech communities. If we consider the legions of people who are affected
by any or all of these circumstances listed above, the amount of bi- or
multilingualism is truly extraordinary. But even if a region is not divided
up into communities speaking syntactically rather different dialects or
variants of a language, educational policy requires the teaching of one or
more foreign languages as part of the curriculum from primary school
onwards or as early as kindergarten. Therefore, even in regions where the
overwhelming majority acquire and speak the same language, the use of

Multi-competence and syntax 207
at least one other language different from the official one has to be
necessarily regarded as the most common and thus standard norm, or as
Cook and Newson express it from a multi-competence perspective: ‘most
people, or indeed all people, have multiple grammars in their minds’
(Cook and Newson 2007, p. 223).
As the present volume suggests, multi-competence is not restricted to
the knowledge of syntax in the mind of multilinguals; nevertheless,
the growth of syntactic knowledge in acquisition plays an undoubtedly
leading role in language development. The role of syntax in language
development is precisely what we try to explore in this chapter. In doing
so we present some essential research findings to claim that a cumulative
enhancement of linguistic knowledge facilitates language acquisition, as
proposed by Flynn et al. (2004), and dwell upon the syntactic implications
such a proposition has for the Faculty of Language and related processes.
The chapter intends to show that the proficient use of multiple languages
enhances learners’ ability to develop syntactic fluency which, we argue, is to
be understood as a certain syntactic sensitivity in detecting target-specific
feature setups for functional categories and combinatorial ease in integrat-
ing and mapping them into the new grammar.
10.2 Development of language and cognitive

development
Most experts agree that people’s capability to acquire languages – the

Faculty of Language (FL) – is specific to the human species, and that this
capacity is innate. The theory of such a genetic endowment has been called
Universal Grammar, or UG, to describe how the human brain processes
linguistic material (see e.g. Chomsky 1986, Wunderlich 2007, or Pietroski
and Crain 2012; for an overview of L1 language development from this
perspective, see Lust 2006). There are other recent but sharply emerging
theories that invalidate the existence of a biologically programmed FL
(see Li and Zhao 2013 for an overview of connectionist proposals for
cognitive architecture, and de Bot, Lowie and Verspoor 2007 for a descrip-
tion of Dynamic Systems Theory). It is not our intention here to debate
the convenience of conducting the study of syntax within the generative
tradition, the series of crucial theoretical assumptions which have proved
most fruitful in the last 50 years, but, aided by the assumption of an
innate capacity for language, we believe we may succeed in tracing
some basic components for the theory of language development which
considers bi- or multilingualism as Nature’s standard.
It has always proved to be a puzzling matter to discover the fundamen-
tal nature of the human innate capacity for language; hence language
acquisition has always evoked considerable interest among linguists.
From the perspective of syntax, there are two essential elements that

208 É V A B E R K E S A N D S U Z A N N E F LY N N
fundamentally determine how linguistic theory, and by extension the

theory of mind, are to be developed. Researchers are divided about
the stand they take in these matters. First, linguists are interested in
finding out how the development of language differs from other cogni-
tive development and thus the relationship between the FL and other
domains of cognition. Can language development be considered as one
of the factors of a more general cognitive development governed by
external rules or does it have its own specific principles which govern
linguistic development in general? Once the proposal of an independent
FL for language development is accepted, a corollary question arises as to
its availability for L2. . .Ln acquisition. Empirically founded arguments
and theoretically valid considerations have been made to support the
idea of an innately determined faculty specific to human language that
guides and constrains L1 development as well as later L2–Ln acquisition
(see e.g. Epstein et al. 1996 and Martohardjono et al. 1998, among others).
This does not mean, however, that the theory of an independent FL
denies that language acquisition is a rather complex phenomenon,
subject to and influenced by the development of other related cognitive
systems. Linguistic theory must focus on what is distinctive and thus
specific to language and language acquisition; only then may it contri-
bute to the construction of an adequate theory of mind.
As an illustration of how the access to unknown linguistic phenomena
can be studied by L2 learners, consider Ko, Perovic, Ionin and Wexler’s
(2008) study of how certain semantic universals affect the choice of
English L2 articles. The authors of the study tested two groups of English
L2 learners on their knowledge of three semantic factors in English L2:
definiteness, specificity and partitivity. The learners’ L1s, Serbo-Croatian
for group 1 and Korean for group 2, lack articles and demonstrate no other
way of directly encoding either specificity or partitivity. Nonetheless,
results show that learners with either of the L1s can distinguish specificity
in English L2. Since the learners could not rely on learnt knowledge taken
from their L1 in either of the cases, these results cannot be explained by
language transfer. Hence, this finding presents support for the claim that
learners of English L2 can have direct access to linguistic knowledge
provided by their UG without the intermediation of their L1.
Another fundamental question related to the FL refers to its change
over time. More specifically, language theory must ask what the role of the
FL is in language development. What happens to the FL in the course of
development, whether L1, L2 or Ln? Does the FL alter at all, and if so, how can
alteration in the FL be conceived or understood? Accordingly, the ultimate
question refers to the nature of language development, namely to ask what it
is that matures, changes or restructures as a response to input effectuating
increased knowledge of language. If we follow the line of investigation we
have adopted here, the logical answer to these questions is that the FL cannot
change, mature or restructure at all, being an innately determined language

capacity for humans. Nevertheless, the increase in knowledge of language

takes place as a natural consequence to linguistic input. To give an example,
let us consider here a behavioural acquisition study comparing the structural
constraints of children and adult code-switching. Paradis, Nicoladis and
Genesee (2000) looked at how production data of child and adult French–
English bilinguals reflect learners’ knowledge of the functional category
INFL for these languages. The presence of system morphemes in their utter-
ances, such as tense, agreement or aspect markers, copulas, or auxiliaries, is
considered as overt reflexes of the above-mentioned functional category.
Given the fact that INFL-related morphology in English emerges later
in development than in French, results indicate that the use of language-
specific INFL morphology does not necessarily reflect children’s syntactic
knowledge related to the language-specific INFL. The authors conclude that
syntactic knowledge must be regarded as something different from the
knowledge of grammatical rules of the target language. Yet, syntactic knowl-
edge seems to be closely related to the acquisition of the target-specific setup
of elements in the Lexicon, in this case the specific architecture of the
English and French functional category INFL.
10.3 Bilingualism is Nature’s standard
In Section 10.2 we introduced the two fundamental questions a syntactic

theory of language development must consider if it aims to be explanatory.
In what follows, we would like to delineate an approach which may
eventually help to understand language development of multilinguals.
Here we not only present supporting evidence from a wide range of
behavioural L1 and L2–Ln acquisition studies but also provide arguments
using the results of some recent brain imaging studies.
Linguistic theory, especially since the formulation of the framework of
Principles and Parameters (Chomsky 1981), has stimulated fruitful
research in language acquisition. A great variety of language combinations
have been studied, especially those where parameter values do not match.
The term parameter resetting refers to the mechanism which helps learners
to assign a new value to the learnt language. The term itself seems to be a
rather unfortunate choice as it suggests the loss of a value with the
replacement of another, which would ultimately imply the loss of the
representation of one language-specific grammar, viz. L1, in the acquisi-
tion of another, viz. L2. Not only is it a problematic term, but practical
considerations heavily undermine the explanatory aspirations of a theory
that adopts ‘parameter resetting’. Individuals with linguistic competence
in at least two languages necessarily own and retain more than one
grammar; the simultaneous presence of different grammars in a mind is
a fact (see explicit examples in Cook and Newson 2007). Simultaneous
bilinguals, for instance, are in multiple language states at the same time.

Given the great diversity in how languages generate wh-questions,

wh-movement has been one of the most studied areas in language theory
and acquisition. Let us take a Korean–English simultaneous bilingual as an
example. It is well known that the two languages represent typologically
contrasting approaches. Whereas English requires that wh-elements
move to the C-domain visibly in the course of the derivation (1a), Korean
wh-phrases remain in situ (1b):
(1) a. Whati do you like ei?

b. ne-nun mwues-ul cohaha-ni?
you-Top what-Acc like-QE
(What do you like?)
(Chung 2000, p. 358)
Simultaneous bilinguals, i.e. those who learnt both languages from birth,
are speakers with two L1s, hence their FL is assumed to represent the two
languages in fundamentally similar ways. The central issue here is to find a
plausible way to reconcile linguistic theory with the theory of language
acquisition, theory with practice, theory with how multiple linguistic
systems actually interact in the mind of an individual. More specifically,
if we accept that wh-movement is triggered by a strong wh-operator feature
on the head of a functional category like CP (Chomsky 1995), the question
is how a person who is in English and Korean language states at the same
time can select the correct strong/weak wh-operator for an interrogative
sentence. Consequently, and as indicated above, the technical term of
parameter setting/resetting is not able to capture the important aspect
we have dwelled on here.
Recent developments in code-switching studies further elucidate this
intriguing question. MacSwan (1999, 2000, 2013) convincingly argues for
the implementation of a constraint-free theory of code-switching, i.e. a
theory without a speculative control structure whose role would be to
mediate between the two grammars in use. Making effective use of the
tools provided by the Minimalist Program (Chomsky 1995), MacSwan
proposes that cross-linguistic variation can be traced back to the Lexicon,
whereas the syntactic operations along the computation are constrained
by UG for all intervening languages. Cheng’s Clausal Typing Hypothesis
(Cheng 1991) explains how the typological variation in wh-question forma-
tion can naturally be applied to the Lexicon. In order to ‘type’ a clause as
interrogative, the speaker either selects a wh-particle for the C-head, in
which case overt movement is not necessary, or chooses a wh-word, which
has to be valued by an AGREE operation according to the requirements of
the overt component, and movement follows. If this hypothesis is on the
right track, an English–Korean multilingual speaker selects the interroga-
tive pronoun what from the Lexicon in English and, as required by UG,
moves it to the C-domain to construct the sentence (1a), but chooses the

wh-particle -ni and leaves it in situ to express the same idea in Korean (1b).
Further support for the view that cross-linguistic differences are restricted
to lexical representations is provided by insights about the development of
the functional INFL category. As noted above, Paradis, Nicoladis and
Genesee (2000) investigated code-switching data from 15 bilingual chil-
dren and found an unequal emergence of the INFL category in the acquisi-
tion of the two languages. The authors claim that English–French bilingual
children, although clearly aware of this category’s existence, seem to lag
behind in their productive use of INFL in English. Such a finding hints at
the possibility of regarding the feature setup of the English INFL as less
transparent for children, which effectuates such a delay. Further studies
are necessary to make a stronger claim for the lexicalist approach
proposed by MacSwan, but the hypothesis has already proved to be
productive, and if it continues doing so, it undoubtedly has far-reaching
implications, not only for how the architecture of the bilingual FL may
look but also for the theory of language acquisition in general.
10.4 Towards a new research paradigm for language

development
Making the statement that bilingualism is Nature’s standard strongly

implies that an efficacious theory of language development must show a
serious interest in trying to understand the bilingual mind. In what fol-
lows, we would like to suggest some approaches to constructing a new
paradigm for the theory of language development, pointing out how they
are related to the two fundamental questions of language acquisition
discussed above, in order to improve our comprehension of what actually
happens in the field of syntactic development. We believe that such an
endeavour will also foster understanding of related domains of human
cognition.
10.4.1 Aspects of grammatical knowledge development

In the present section we would like to discuss some relevant issues
regarding what aspects of grammatical knowledge develop in the
course of language acquisition. However, before we set off, we need
to say some words about what we consider to be the ‘initial state’ for L2
acquisition. There are two principal and rather contrasting hypotheses
on which researchers within the UG tradition have based their models
for language development: the Maturation Hypothesis, originally put
forward by Felix (1984) and the Strong Continuity Hypothesis, proposed
by Flynn and Lust (2002). (A comparable model for L1 development was
originally proposed by Levy 1983, challenging Gleitman’s 1981 stage
model for language development.) The difference between the two

comes down to the fact that the Maturation Hypothesis assumes L2

learners’ access to UG in the form of a previously developed language-
specific grammar, namely their L1, whereas the Strong Continuity
Hypothesis posits that UG remains intact and available for guiding
the development of further target-specific grammars. Adopting one or
the other hypothesis has far-reaching implications for a theory of
language development. The Maturity Hypothesis only allows for a
unidirectional or linear development, whether increasing or decreas-
ing, viz. language loss, due to UG having been transformed into L1. It
seems that such a hypothesis is doomed inoperable for a research
paradigm that argues for bilingualism as Nature’s standard. The
Strong Continuity Hypothesis, on the contrary, allows for postulating
an initial state for L2–Ln acquisition similar to that for L1 regarding the
capacity of a speaker’s FL to construct target-specific grammars. In
other words, it proposes that certain principles of UG operate as long
as a learner encounters new linguistic data that need to be processed;
see Berkes and Flynn (2013a) for a more detailed review of the two
hypotheses and the models based on them.
Once we accept the claim that, after shaping L1, UG remains accessible
for the processing of subsequent linguistic data, we need to ask why the
majority of L2 learners seem to fall behind L1 speakers in their perfor-
mance of the given language. The Critical Period Hypothesis has been the
subject of intense debate for decades. The original idea proposed by
Penfield and Roberts (1959) and broadly disseminated by Lenneberg
(1967) was also extended to SLA; for detailed reviews of this debate see
Singleton and Lengyel (1995) or Singleton and Ryan (2004). Nonetheless,
research into behavioural L2 acquisition indicates that, contrary to popu-
lar belief, there is no empirical evidence in support of a critical period for
L2 (or next language) acquisition (e.g. Epstein et al. 1996; Klein and
Martohardjono 1999; Cook 2002b). In addition, several brain imaging
studies using fMRI and PET appear to corroborate this claim by revealing
that highly proficient non-native speakers display similar patterns of acqui-
sition, regardless of age of acquisition (Perani et al. 1998).1 This is in contrast
to the results produced by intermediate-level non-native speakers, who dis-
play a more diffuse representation of the L2 across the brain when compared
to highly proficient non-native speakers (Dehaene et al. 1997). The same
conclusion was drawn by Kessler, Martohardjono and Shafer (2004), who
also looked at morphosyntactic development to test abstract grammatical
knowledge. In short, language-related brain activation patterns seem to
attest that it is language proficiency rather than age of acquisition that is
critical in determining how the brain processes linguistic material.
Development in applied electroencephalography (EEG) studies
further attest that event-related potential (ERP) responses vary system-
atically with L2 proficiency level. Increased proficiency seems to be in a
causal relationship with late P600 effects, viz., 600 milliseconds after

the introduction of the stimulus to the brain. Nonetheless, results also

indicate omissions of left anterior negativity. This in turn suggests less
automated syntactic processing in the case of advanced learners (Hahne
2001). Magnetoencephalography (MEG), a further functional neuroima-
ging technique, was employed to conduct a recent study to map brain
activity in language production (Arviso et al. 2005). The main objectives
of this study were to learn how MEG patterns of activation differ in the
course of L2 development and whether patterns of L2 acquisition can in
any way, at any stage, be related to established L1 patterns. Results
indicate that learners of English as an L2 differentiate between gram-
matical and ungrammatical verb stimuli in the same manner as native
speakers. Consistent with former results in behavioural SLA studies,
the MEG study of Arviso et al. also suggests that age of arrival and age
of learning an L2 are not significant factors in determining brain
patterns for language acquisition. We conclude then that any theory
of language development must clearly recognize and distinguish
between factors of performance and those of competence, taking pro-
ficiency rather than age as a covariate. Having said this, we now return
to the issue of how languages known prior to a new acquisition task
relate to each other. This has been the subject of heated debates over
the years, either in the form of L1 transfer, to refer to how the L1 of a
learner influences further acquisition, or as cross-linguistic influence, to
refer to how any or all languages known prior to the acquisition of an
Ln impact this process. For recent discussions see, for example, De
Angelis (2007), Hammarberg (2009) or Leung (2009); for an up-to-date
review see Berkes and Flynn (2013a).
As we indicated above, the Strong Continuity Hypothesis suggests that
UG is available in its entirety to the learner, yet accepting this hypothesis
does not imply that the role played by previously acquired languages and
language grammars in subsequent language development must be
rejected. More precisely, what is predicted is that language learners do
not ‘transfer’; that is, they do not build the new grammar using the
language-specific grammatical features of the L1, but rather they construct
the new grammar with the help of ‘grammatical mapping’, i.e. they ‘map
from one primary structure to a more developed structure by dissociating
modular grammatical components and integrating them in the “assem-
bly” of new language-specific grammars’ (Flynn et al. 2005, p. 2). All in
all, here we propose to reject ‘transfer’ on the level of fundamental
computational mechanisms which characterize the FL as a result of
understanding language learning, not as a maturational, but rather a
computational process.
Such an understanding obviously raises the question of how previously
known languages may be at work when a learner faces a new acquisition
task. Results from a recently completed series of studies may shed light on
this matter (Flynn and Lust 1980; Flynn 1983, 1989; Flynn et al. 2004;

Berkes and Flynn 2012a, 2012b). Learners with different L1s were tested on
their elicited imitation of three types of restrictive relative clauses, headed
(specified and unspecified) and free relatives, each type including four
variants according to function of head NP and gap (SS, SO, OS, OO), as
illustrated in (2)–(5) below:
(2) Headed (specified): The woman instructed the lawyer who the
policeman called. (OO)
(3) Headed (unspecified): The person who the engineer answered

criticized the man. (SO)
(4) Free: Whoever entered the office introduced the professor. (SS)
(5) Free: The professor introduced whoever greeted the lawyer. (OS)
The participating homogeneous groups consisted of English L1 children,

or either English L2 or English L3 learners at three levels of English
proficiency, measured by a standardized test prior to experiment.
Background languages for the groups were selected according to the para-
metric variation of two CP-related features: head-directionality and word
order in subordinate clauses. Table 10.1 lists the participating groups and
their background languages, indicating the relevant characteristics with
regard to the two CP features in each language.
The general results of this series indicate, on the one hand, that a linear
model for multiple language acquisition does not satisfactorily explain
the development of underlying syntactic processes. On the other hand,
the experiments show that the development of syntactic knowledge in L3
acquisition cannot be clearly traced back to either the L1 or to the
influence of the last learned language. Japanese L1 learners of English
L2 (study #2 in Table 10.1), like the English monolingual children (study
#1 in Table 10.1), scored significantly higher on free relatives than on any
of the lexically headed relative types, whereas the Spanish L1 group
(study #3 in Table 10.1) did not do significantly better on any of the
three types of relatives in their L2 acquisition of English, despite having
been equated at all levels of English competence. It seems that the
Spanish L1 learners could somehow draw upon their knowledge of CP
structure, for Spanish and English match in terms of both CP properties
(contrary to the Japanese L1 learners of L2 English), and use it in subse-
quent learning. This seems to signify that the free relative clause struc-
ture is a developmental precursor to the lexically headed form (similar
results were found for English, Hamburger 1980 and Flynn and Lust 1980;
for Mandarin, Packard 1988; for Korean, Lee 1991 and Lee et al. 1990; for
Japanese, Murasugi 1991; for French, Foley 1996; for Tulu, Somashekar
1999; for Polish, Mróz 2010).
The primacy of free relatives in CP development of the target language
was further tested by Flynn et al. (2004) (study #4 in Table 10.1), who

Table 10.1 CP-related features in background language, from selected

studies
Head- Head-
# Study Source Language initial final SVO SOV
1 English L1 Flynn and Lust (1980) English ✓ ✓

2 JaL1/EnL2 Flynn (1989) Japanese ✓ ✓
3 SpaL1/EnL2 Flynn (1983, 1989) Spanish ✓ ✓
4 KazL1/RuL2/ Flynn et al. (2004) Kazakh ✓ ✓
EnL3
Flynn et al. (2004) Russian ✓ ✓
5 HuL1/EnL2 Berkes and Flynn (2012b) Hungarian ✓ ✓
6 HuL1/GeL2/ Berkes and Flynn (2012a, German ✓ ✓
EnL3 2012b)
investigated an adult group of KazL1/RuL2 learners of L3 English. Kazakh is

a head-final language with an SOV constituent word order within the
relative clause, like Japanese, the difference being that these learners
learned Russian as L2 first and subsequently English as L3; Russian is a
head-initial language with primary SVO order in relative clauses, so it
matches the English configuration. Results of this study indicated that
the development of the CP structures in the target language of these
learners patterned with that of the SpaL1/EnL2 rather than the JaL1/EnL2
group; in other words, the Kazakh L1 adult learners, with the help of their
experience in acquiring L2 Russian, had developed a target-like CP struc-
ture upon which they could draw. Flynn et al. (2004) concluded on the one
hand that the universal knowledge underlying the free relative seems to be
fully available for the learner and, thus, acts as a developmental precursor
in the acquisition of English, as was to be observed in the case of the JaL1/
EnL2 group; on the other hand, prior CP development appears to influence
the development of target-specific CP structure, as in the case of the KazL1/
RuL2/EnL3 group or the Spanish L1 learners of English, who had already
instantiated the CP setup for English through the acquisition of their own
L1. In sum, it seems that some sort of awareness of CP-related features in
previously known languages has an impact on how additional language-
specific grammars develop. Based on these experimental facts Flynn et al.
(2004) ended up postulating the Cumulative Enhancement Model for Language
Acquisition (CEM), which claims that learners may draw upon acquired
knowledge of language-specific information or template in later acquisi-
tion, e.g. CP features. In other words, this model hypothesizes that all
previously known languages are available to the learner to constructively
enhance subsequent language learning.
Berkes and Flynn (2012b) extended the series by adding production data
from a bilingual Hungarian L1/English L2 group and a multilingual
Hungarian L1/German L2/ English L3 group to the database. Hungarian

and German match English in head-directionality, notwithstanding that

both Hungarian and German differ from English in another CP-related
parameter, the word order in the subordinate clause, as indicated in
Table 10.1. German exhibits an SOV word order in subordinate relative
clauses, whereas Hungarian is considered to be organized around the
concept of topic and focus rather than the concept of subject (Li 1976),
which also suggests that constituent word order within the relative clause
might not be encoded in this language the same way as in German or in
English. Results of study #5 reveal that free relatives were significantly
more productive than the lexically headed forms in terms of amount
correct, i.e. the HuL1/EnL2 learners appear to develop the target-like
CP structure over time, whereas the free relatives, as in the case of the
JaL1/EnL2 group (study #2 in Table 10.1), seem to be fully available at all
points in development. Free relatives seem to have a more transparent
syntactic structure for learners (Flynn et al. 2005); subsequently they
appear to be available for them from the beginning of their language
development. Hence, this study provides further support for the claim
that free relatives seem to be developmental precursors to full lexically
headed forms. In contrast, the multilingual HuL1/GeL2/EnL3 speakers
produced all types of relative sentences with nearly the same high
rate of success. The additional error analyses showed that the emerging
developmental patterns were radically distinct in the English L2 and L3
groups. No significant difference was detected in their production of free
relatives at any stage of their language development. It appears then that
multilingual language growth regarding syntactic development shows a
highly nuanced picture. Multilingual learners of English exhibited an
enhanced knowledge of the target-specific CP architecture, although
German does not entirely replicate the target-specific CP. In this case
learners could not possibly draw on already established patterns, viz.
previously acquired complete CP structure, nevertheless a significant
enhancement effect was found, as if the previously acquired fully fledged
CP structure of L2 German had activated possible CP patterns in the
learners’ mind.
The series of studies presented here provide a counter-example to
simplified accounts of language transfer from either L1 or any previous
language. Any theory of acquisition based entirely on transferring surface
elements from one language to another cannot give an explanatorily
adequate account for how language develops in the mind of the learner.
There are, however, indices which show that specific previous linguistic
knowledge does make a difference in subsequent language development.
Not only knowledge of a fully fledged CP structure may induce learners to
skip a more ‘primitive’ stage and build up complex structures in the target
language, as we saw in the case of L1 Kazakh learners of L3 English who
could draw upon their L2 Russian (study #4 in Table 10.1), but also the
activation for new CP patterns in the mind of multilinguals (study #6 in

Table 10.1). The subtle knowledge that CP may have diverse feature setups
seems to provide syntactic fluency, the ability to combine features for CP
with greater ease to facilitate subsequent acquisition of target-specific
grammars and circumvent falling back onto the more primitive stage of
building upon free relatives.
10.4.2 What knowledge does the FL provide?

In Section 10.4.1 we attempted to show that multiple language acquisition
is to be considered fundamentally different from L2 acquisition with
regard to the linguistic properties of the initial point of the non-native
grammar. Due to acquired syntactic knowledge provided by previously
known languages, the learner of an Ln seems to be equipped with the
ability to dissociate and newly integrate functional features for elements
selected from the Lexicon to compute sentences, a disposition which may
result in a statistically measurable cumulative enhancement effect con-
cerning the new grammar. Now we turn to considering the type of knowl-
edge a multilingual learner’s FL provides, or in other words, we undertake
to explore how learners integrate innate linguistic and language-specific
knowledge for Ln in acquisition.
Some mainstream acquisition studies brought forth the claim that lan-
guage development is mainly due to non-linguistic mechanisms, such as
distance and processing ease. In point of fact behavioural studies on
restrictive relative clauses found that there is a general hierarchical
order in ease of processing, equally applicable to comprehension
and production data in L2 English acquisition. According to this view,
subject-modifying relatives, especially where the gap in the relative clause
is the subject (SS) were claimed to be the easiest to process (e.g. Gass 1979;
Pavesi 1986), mainly due to the shorter distance between the filler and the
gap in the relative clause in the linear surface order of constituents (e.g.
Tarallo and Myhill 1983). On the other hand, recent usage-based analyses
have attested the primacy of object-modifying relatives in development
of complex structures in L1 acquisition (e.g. Kidd et al. 2007) where lan-
guage-specific difficulties in processing object-modifying relatives were
eliminated by employing pronominal subjects or inanimate objects. Yet,
abandoning the quest for underlying syntactic principles leaves essential
cross-linguistic second language acquisition findings unexplained.
Therefore, if the theory of dissociation and integration of features is on
the right track, we need to look for empirical support that language
development is guided and constrained by UG, i.e. indices for a syntax-
driven progress in language development. We believe that experimental
data taken from the acquisition of free relatives provide the desired
arguments.
Results of the series of studies we presented in Section 10.4.1 (see
Table 10.1) help to better understand language acquisition, and most

particularly the nature of the hypotheses that language learners entertain

when learning a new target language. In particular, strong support was
found for the claim that the acquisition of complex sentence structures
involving relativization of an NP in a target grammar requires a specific
developmental process. On the one hand this process seems to follow a
common path for learners with diverse language backgrounds, but on the
other hand it appears to be greatly influenced by accumulated knowledge
of possible syntactic structures in general. The series of studies listed in
Table 10.1 shows that in constructing the lexically headed relative clause
in English, L1, L2 and L3 learners build upon knowledge of the free relative.
This suggests that the universal properties underlying the free relative are
fully available for learners at all points in the course of developing the
target-like grammar; therefore we may say that free relatives seem to act as
a sort of syntactic primitive with respect to other relative clause types and
thus provide a basis for ‘grammatical mapping’ from UG principles to
language-specific grammars (Foley 1996).
It is to be noted here that grammatical function primacy regarding free
relatives is to be distinguished from the Accessibility Hierarchy proposed
by Keenan and Comrie (1977). The latter presents a continuum regarding
the level of difficulty with which language learners access the rules of the
target language based on processing difficulties. Given that free relatives
are structurally more transparent, as noted above, grammatical function
primacy in free relatives reveals an aspect of grammar that is more directly
related to basic or ‘core’ grammar rules of UG. Hence free relatives provide
the basis for ‘grammatical mapping’ of complex structures, and this is the
reason why examining their development may give insights into the type
of knowledge the FL provides.
A further study was conducted based on the series listed in Table 10.1 to
investigate this matter. Berkes and Flynn (2013b) compared the elicited
imitation of free relatives of a German L1 and a Hungarian L1 group of
adult English L2 learners. This study involved the four variants of free
relatives according to function of head and gap (SS, SO, OS, OO) at three
levels of language proficiency measured by an independent standardized
test, as presented earlier. The syntactic derivation of an OO free relative
based on assumptions made by Caponigro (2000) proves to be the most
economical, and thus the fact that both groups produced statistical results
for the OO type of free relatives2 gives support to the theory that language
development is constrained by principles of economy. The observed gram-
matical role primacy effects therefore imply that language development is
syntax-driven, and more generally, that acquisition of complex structures
is a true reflection of linguistic economy.
Another supporting argument for a syntax-driven FL representation
comes from a study that examined how multilingual learners acquire
correct referential relations for subject-controlled pronominal anaphora
in adverbial subordination. In Berkes and Flynn (2015) we looked at the

elicited production of sentences involving adverbial subordinate clauses

with the subordinator when (6)–(9) of a group of Hungarian L1/German L2
learners of English L3:
(6) When he entered the office, the janitor questioned the man.
(7) When PROi inspecting the room the workeri questioned the
janitor.
(8) The man answered the boss when he installed the television.
(9) The professori answered the owner when PROi preparing the
lunch.
Testing the production of null anaphora (PRO) was critical to the design, as
null anaphora is obligatorily subject-controlled in English, as indicated
(sentences (7) and (9)). However, this is not the case in either Hungarian or
German, nor are there non-tensed clauses; thus, production of such
constructions reflects a form of grammatical analysis implicitly carried
out by these learners.
Accepting the claim that directionality as set in the target-language
fundamentally affects the way learners interpret pronouns, the experi-
ment tested whether referential relations involving pronoun anaphora
would reflect learners’ subtle syntactic knowledge of English directional-
ity. As we mentioned before, neither German nor Hungarian match the
feature setup of the English CP completely: learners must figure it out in
the acquisition process. Although we found no statistically measurable
evidence that learners imitate postposed adjuncts with overt pronouns
more successfully than preposed ones, error analyses suggest that this is
so. Learners tend to commit significantly more anaphora errors on pre-
posed adjuncts with an overt pronoun than on postposed adjuncts at low
level. This fact shows that there is a surprising directionality effect present
in the imitation data and which replicates the results found in English L1
acquisition where it was seen that children tended to be more productive
on forward pronominalization (see Lust et al. 1986, and references cited
therein). It seems that Hungarian L1/German L2 learners of English L3,
like English L1 children, connect directionality of the target-input and
referential relations since they are more productive on forward
pronominalization.
Our first conclusion then is that the directionality principle, as in the
case of English L1 children, constrains the development of learners’ repre-
sentation of the new grammar from the very beginning. Furthermore,
Hungarian is considered to be a null-subject language (NSL), i.e. it requires
the omission of personal pronouns in neutral contexts. Given the fact
that Hungarian does not encode directionality in the domain of our stimuli
as it lacks the obligatory DP layer present in non-null-subject languages
(non-NSL), we assume it is the German L2 that provides this subtle

knowledge learners could draw upon. Production data in this case sup-
ports the hypothesis that language learning is a cumulative process
whereby learners have access to previous syntactic experience to enhance
subsequent learning, as proposed by the CEM. Language representation is
not redundant: syntactic feature settings can be drawn upon to build
a new target-specific grammar. Therefore, production data of adverbial
pre- and postposed adjunct clauses with overt pronouns and with subject-
controlled pronominal anaphora PRO imply that development of
anaphoric relations in these domains is likewise constrained for
Hungarian L1/German L2 learners of English L3 by their FL as for English
L1 children.
The studies presented in this chapter attempt to illustrate how structure
is imposed upon a developing target grammar, in this case English. Such a
finding has far-reaching consequences for designing theories of language
development. If we are on the right track, parameter setting must be rein-
terpreted to refer to the capacity of multilingual learners to respond to the
structural requirements the FL imposes on the new grammar. This is what
we propose to call syntactic fluency. In sum, we could see that our Hungarian
L1/German L2 learners of English L3 could develop the correct anaphoric
relations based on their capacity to dissociate directionality from CP and
integrate it into DP. Simultaneously, success required the recognition of
the obligatory presence of a DP layer in the domain of our stimuli.
10.5 Conclusions
As suggested in this chapter, the view presented by multi-competence as

‘the knowledge of more than one language in the same mind’ (Cook 2012)
necessarily includes the L2. . .Ln speakers’ syntactic knowledge as an inte-
gral part of their language awareness. The basic claim of this paper is
precisely to support the thesis that bi- or multilingualism is Nature’s
standard, which has clear repercussions in the way syntactic development
of language occurs in the brain of language learners. A bilingual mind, a
truly unique human characteristic, is capable of being in multiple lan-
guage states at the same time. This implies that language theory and the
theory of language development must undergo a change in the research
paradigm regarding fundamental issues. One such matter is the way in
which language influence is to be understood, or more generally, the
relationship among known languages is to be considered. We claim that
the unidirectional approach has to be reinterpreted to allow for the more
creative progress the learner goes through. The same is true for how
language parameters and parameter setting must be treated. If features
on lexical and functional elements determine how syntactic relationships
are formed and meaning processed, language development must be the
result of how learners dissociate and integrate lexical and functional

features into elements of the Lexicon to map the linguistic material to

language-specific grammars. Generally speaking, all language learners
must be considered as linguists, albeit unconscious, whose language learn-
ing process is fundamentally determined by how successful they are in
discovering and assigning adequate features to elements picked out from
the Lexicon of the developing target-language. This is one of the aspects of
what we called syntactic fluency, the ability which seems to be enhanced by
linguistic experience provided by knowledge of multiple languages.
Furthermore, there is enough supporting evidence to claim that the
mind imposes structure on developing grammars, which implies that
mapping of linguistic material to grammar representations by learners is
achieved with the help of genetically available support attributed to UG,
and the syntactic requirements of language-specific grammars. Multi-com-
petent plurilingual learners who show high proficiency in languages
where certain feature settings do not match appear to be sensitive at
singling out the features needed for a new setup, another aspect of their
sophisticated syntactic fluency. The analysis of some key issues looked at in
this chapter provides some new insights into a more nuanced understand-
ing of the human capacity for language. We believe that by setting out
from here we have a steadfast chance of developing an explanatory theory
of language development.
Notes
1. Kim et al. (1997) indeed found that age of acquisition emerged as a

significant factor in terms of subsequent patterns of brain activa-
tion. However, their study did not control for levels of L2
competence.
2. Group-specific differences were explicated by the fact that German,
Hungarian and English differ in the ability of the matrix verb in relative
clauses to assign structural case. Matching effects occur in German but
not in Hungarian and only partially in English. Nonetheless, this study
shows that language development, as exhibited in the case of free
relatives, is comparable regardless of the L1.
References
Adams, J. N. and Swain, S. 2002. ‘Introduction’, in J. N. Adams, M. Janse and

S. Swain (eds), Bilingualism in Ancient Society: Language Contact and the
Written Word, pp. 1–20. Oxford University Press.
Benjamins.

Arviso, D., Cuervo, C. M., Flynn, S., Jacobs, K., Lai, K. and Schulz, M. 2005.
MEG evidence that age of second language learning does not matter:
MEG results isolate ELAN for both L1 and advanced L2 speakers of
English. Paper presented at Annual Meeting of the Linguistic Society of
America, Oakland, CA.
Auer, P. and Li Wei. 2007. ‘Introduction: Multilingualism as a problem?
Monolingualism as a problem?’, in P. Auer and Li Wei (eds), Handbook of
Multilingualism and Multilingual Communication, pp. 1–14. Berlin: Mouton
De Gruyter.
Berkes, É. and Flynn, S. 2012a, ‘Further evidence in support of the
Cumulative-Enhancement Model: CP structure development’, in
J. Cabrelli Amaro, S. Flynn and J. Rothman (eds), Third Language
Acquisition in Adulthood, pp. 143–164. Philadelphia and Amsterdam: John
Benjamins.
Berkes, É. and Flynn, S. 2012b. ‘Enhanced L3. . .Ln acquisition and its
implications for language teaching’, in D. Gabrys-Barker (ed.), Cross-
linguistic Influences in Multilingual Language Acquisition, pp. 1–22. Berlin
and Heidelberg: Springer.
Berkes, É. and Flynn, S. 2013a. ‘Multilingualism: new perspectives on
syntactic development’, in T. K. Bhatia and W. C. Ritchie (eds),
pp. 137–167.
Berkes, É. and Flynn, S. 2013b. ‘Economical’ English L2 acquisition by L1
German and L1 Hungarian learners: mapping of grammatical functions
in free relatives provides for structure dependent learning. Paper pre-
sented at ISB9, Singapore.
Berkes, É. and Flynn, S. 2015. ‘Multilingual acquisition of English: devel-
opment of grammar through study of null anaphora’, in M. P. S. Jordà
and L. P. Falomir (eds), Learning and Using Multiple Languages: Current
Findings from Research on Multilingualism, pp. 112–133. Newcastle upon
Tyne: Cambridge Scholars Publishing.
Bhatia, T. K. and Ritchie, W. C. (eds) 2013. The Handbook of Bilingualism and
Multilingualism, second edition. Chichester, UK: Wiley-Blackwell.
Caponigro, I. 2000. ‘Free relatives as DPs with a silent D and a CP comple-
ment’, in V. Samiian (ed.), Proceedings of the Western Conference on
Linguistics, vol. 12, pp. 140–150. Fresno, CA: Department of Linguistics,
California State University.
Cheng, L. L. S. 1991. On the Typology of Wh-Questions. PhD dissertation, MIT
Press.
Chomsky, N. 1981. Lectures on Government and Binding. Dordrecht,
Netherlands: Foris.
Chomsky, N. 1986. Knowledge of Language : Its Nature, Origins, and Use. New
York: Praeger.
Chomsky, N. 1995. The Minimalist Program. Boston, MA: MIT Press.
Chomsky, N. 2006. The Architecture of Language. New Delhi: Oxford
University Press.

Chung, D. 2000. ‘On the representation and operation of WH-questions’,

The Korean Generative Grammar Circle: Studies in Generative Grammar 10,
357–387.
Cook, V. J. (ed.) 2002a. Portraits of the L2 User. Clevedon: Multilingual
Matters.
Cook, V. J. 2002b. ‘Background to the L2 user’, in V. J. Cook (ed.), pp. 1–28.
Cook, V. J. and Newson, M. 2007. Chomsky’s Universal Grammar: An
Introduction, third edition. Oxford: Blackwell.
Dehaene, S., Dupoux, E., Mehler, J., Cohen, L., Paulesu, E., Perani, D., van de
Moortele, P.-P., Lehéricy, S. and Le Bihan, D. 1997. ‘Anatomical variabil-
ity in the cortical representation of first and second languages’,
Neuroreport 8, 3809–3815.
Epstein, S. D., Flynn, S. and Martohardjono, G. 1996. ‘Second language
acquisition: theoretical and experimental issues in contemporary
research’, Behavioral and Brain Sciences 1, 677–758.
Felix, S. W. 1984. Two Problems of Language Acquisition: On the Interaction of
Universal Grammar and Language Growth. Passau: Universität Passau.
Flynn, S. 1983. A Study of the Effects of Principal Branching Direction in Second
Language Acquisition: The Generalization of a Parameter of Universal Grammar
from First to Second Language Acquisition. Dissertation, Cornell University,
Ithaca, NY.
Flynn, S. 1989. ‘Spanish, Japanese and Chinese speakers’ acquisition of
English relative clauses: new evidence for the head-direction para-
meter’, in K. Hyltenstam and L. K. Obler (eds), Bilingualism across the
Lifespan: Aspects of Acquisition, Maturity, and Loss, pp. 116–131. Cambridge
University Press.
Flynn, S., Foley, F., Gair, J. and Lust, B. 2005. Developmental primacy of free
relatives in first, second and third language acquisition: implications for
their syntax and semantics. Paper presented at Linguistic Association of
Great Britain, Cambridge University.
Flynn, S., Foley, C. and Vinnitskaya, I. 2004. ‘The Cumulative-
Enhancement Model for language acquisition: comparing adults’ and
children’s patterns of development’, International Journal of Multilingualism
1 (1), 3–17.
Flynn, S. and Lust, B. 1980. ‘Acquisition of relative clauses in English:
developmental changes in their heads’, in W. Harbert and
J. Herschensohn (eds), Cornell Working Papers in Linguistics 1, pp. 33–42.

Ithaca, NY: Department of Modern Languages and Linguistics, Cornell

University.
Flynn, S. and Lust, B. 2002. ‘A minimalist approach to L2 solves a dilemma
of UG’, in V. J. Cook (ed.), pp. 95–120.
Foley, C. 1996. Knowledge of the Syntax of Operators in the Initial State: The
Acquisition of Relative Clauses in French and English. Dissertation, Cornell
University, Ithaca, NY.
Gass, S. M. 1979. ‘Language transfer and universal grammatical relations’,
Gleitman, L. 1981. ‘Maturational determinants of language learning’,
Cognition 10 (1), 103–114.
issues’, in T. K. Bhatia and W. C. Ritchie (eds), The Handbook of
Bilingualism, pp. 32–63. Oxford: Blackwell.
Hahne, A. 2001. ‘What’s different in second-language processing?
Evidence from event-related brain potentials’, Journal of Psycholinguistic
Research 30, 251–266.
Hamburger, H. 1980. ‘A deletion ahead of its time’, Cognition 8, 389–416.
Hammarberg, B. (ed.) 2009. Processes in Third Language Acquisition. Edinburgh
University Press.
Keenan, E. L. and B. Comrie. 1977. ‘Noun phrase accessibility and universal
grammar’, Linguistic Inquiry 8 (1), 63–99.
Kessler, K., Martohardjono, G. and Shafer, V. 2004. ‘ERP Correlates of age
and proficiency in L2 processing of syntactic and inflectional informa-
tion’, in A. Brugos, L. Micciulla and E. C. Smith (eds), Proceedings of the
Annual Boston University Conference on Language Development, vol. 28,
pp. 294–305. Somerville, MA: Cascadilla Press.
Kidd, E., Brandt, S., Lieven, E. and Tomasello, M. 2007. ‘Object relatives
made easy: a cross-linguistic comparison of the constraints influencing
young children’s processing of relative clauses’, Language and Cognitive
Processes 22 (6), 860–897.
Kim, K. H. S., Relkin, N. R., Lee, K.-M. and Hirsch, J. 1997. ‘Distinct cortical
Klein, E. C. and Martohardjono, G. 1999. The Development of Second Language
Grammars: A Generative Approach. Amsterdam: John Benjamins.
Ko, H., Perovic, A., Ionin, T. and Wexler, K. 2008. ‘Semantic universals and
variation in L2 article choice’, in R. Slabakova, J. Rothman and
E. Gavruseva (eds), Proceedings of the 9th Generative Approaches to Second
Language Acquisition Conference (GASLA 2007), pp. 118–129. Somerville, MA:
Cascadilla Proceedings Project.
Lee, K.-O. 1991. On the First Language Acquisition of Relative Clauses in Korean: The
Universal Structure of COMP. Dissertation, Cornell University, Ithaca, NY.
Lee, K.-O., Lust, B. and Whitman, J. 1990. ‘On functional categories in
Korean: a study of the first language acquisition of Korean relative

clauses’, in E.-J. Baek (ed.), Papers from the Seventh International Conference on
Korean Linguistics, pp. 312–333. University of Toronto Press.
Lenneberg, E. H. 1967. Biological Foundations of Language. New York: Wiley.
Leung, Y.-K. I. (ed.) 2009. Third Language Acquisition and Universal Grammar.
Levy, Y. 1983. ‘It’s frogs all the way down’, Cognition 15, 75–93.
Li, C. N. (ed.) 1976. Subject and Topic. New York: Academic Press.
Li, P. and Zhao, X. 2013. ‘Connectionist models of second language acquisi-
tion’, in M. del Pilar Garcia Mayo, M. Junkal, G. Mangado and M. M.
Adrian (eds), Contemporary Approaches to Second Language Acquisition,
pp. 177–198. Amsterdam: John Benjamins.
Lust, B. 2006. Child Language: Acquisition and Growth. Cambridge University
Press.
Lust, B. et al. 1986. ‘A comparison of null and pronoun anaphora in first
language acquisition’, in B. Lust (ed.), Studies in the Acquisition of Anaphora,
vol. I: Defining the Constraints, pp. 245–277. Dordrecht: D. Reidel.
MacSwan, J. 1999. A Minimalist Approach to Intrasentential Code Switching. New
York, NY: Routledge.
MacSwan, J. 2000. ‘The architecture of the bilingual language faculty:
evidence from intrasentential code switching’, Bilingualism: Language
and Cognition 3, 37–54.
MacSwan, J. 2013. ‘Code-switching and grammatical theory’, in T. K. Bhatia
and W. C. Ritchie (eds), pp. 323–350.
Martohardjono, G., Epstein, S. D. and Flynn, S. 1998. ‘Universal Grammar:
hypothesis space or grammar selection procedures? Is UG affected by
critical periods?’, Behavioral and Brain Sciences 21, 612–614.
Mróz, M. 2010. Bilingual Language acquisition: Focus on Relative Clauses in Polish
and English. Warsaw: Wydawnictwa Uniwersytetu Warszawskiego.
Murasugi, K. 1991. Noun Phrases in Japanese and English: A Study in Syntax,
Learnability and Acquisition. Dissertation, University of Connecticut,
Storrs.
Packard, J. L. 1988. ‘The first language acquisition of prenominal
modification with de in Mandarin’, Journal of Chinese Linguistics
16 (1), 31–53.
Paradis, J., Nicoladis, E. and Genesee, F. 2000. ‘Early emergence of struc-
tural constraints on code-mixing: evidence from French–English bilin-
gual children’, Bilingualism: Language and Cognition 3, 245–261.
Pavesi, M. G. 1986. A Study of Markedness in Second Language Acquisition. PhD
dissertation, University of Edinburgh.
Penfield, W. and Roberts, L. 1959. Speech and Brain Mechanisms. Princeton,
NJ: Princeton University Press.
Perani, D., Paulesu, E., Sebastian-Galles, N., Dupoux, E., Dehaene, S.,
Bettinardi, V., Cappa, S. F., Fazio, F. and Mehler, J. 1998. ‘The bilingual
brain: proficiency and age of acquisition of the second language’, Brain
121, 1841–1852.

Pew Research Center 2013. Internet and American Life Project. http://pewinter
net.org/Reports/2013/social-networking-sites/Findings.aspx
Pietroski, P. and Crain, S. 2012. ‘The language faculty’, in E. Margolis, R.
Samuels and S. P. Stich (eds), The Oxford Handbook of Philosophy of Cognitive
Science, pp. 361–381. Oxford University Press.
Singleton, D. and Lengyel, Z. (eds) 1995. The Age Factor in Second Language
Acquisition. Clevedon: Multilingual Matters.
Singleton, D. and Ryan, L. (eds) 2004. Language Acquisition: The Age Factor.
Somashekar, S. 1999. Developmental Trends in the Acquisition of Relative Clauses:
Cross-linguistic Experimental Study of Tulu. Dissertation, Cornell University,
Ithaca, NY.
Tarallo, F. and Myhill, J. 1983. ‘Interference and natural language proces-
sing in second language acquisition’, Language Learning 33, 55–76.
Wunderlich, D. 2007. ‘Why assume UG?’, in M. Penke and A. Rosenbach
(eds), What Counts as Evidence in Linguistics: The Case of Innateness, special
issue of Studies in Language, 615–647.

11
Syntactic processing
Leah Roberts
11.1 Introduction
Understanding a sentence in real time, or parsing,1 involves extracting

meaning from each of the incoming constituents, imposing a structural
and semantic analysis on the input, and assessing the plausibility of the
interpretation against both local (bottom-up) and real-world contextual
(top-down) knowledge. Although ubiquitous, parsing is complex, and
these interacting processes occur within milliseconds of encountering
each word. Psycholinguists have been using time-sensitive methods
such as self-paced reading, eye-tracking and electroencephalography
(EEG) to chart these processes in real time for decades, in order to
build a picture of the nature of the human sentence processing
mechanism, but the field of second language (L2) syntactic processing2
is comparatively new. To date, L2 syntactic processing researchers on
the whole work within what Cook (Chapter 1, this volume) calls the
monolingual perspective. That is, a major aim (whether explicitly or
implicitly expressed) of most of those interested in L2 parsing has
been to examine how learners are similar to, and different from, native
monolingual speakers of the target language (TL). This entails the
underlying assumption that monolingual parsing is both the norm
and the ideal. L2 parsing researchers tend to make this assumption,
whether work is undertaken by those trained in linguistics, most of
whom tend to study the syntactic processing of ‘late L2 learners’ (lear-
ners who have first been exposed to the TL after the purported ‘critical
period’: see Lenneberg 1967), or whether the research is undertaken
from a psychological perspective. Therefore, shifting from the mono-
lingual to a multi-competence perspective would be rather a challenge for
the majority of L2 processing researchers.
It is of course unsurprising that the monolingual perspective is taken in
L2 syntactic processing work. This is because until very recently, research

228 LEAH ROBERTS
into the mechanisms that underlie the human sentence processor has
focused almost exclusively on monolingual (most often English) speakers
(Blumenfeld and Marian 2013; van Gompel 2013). Furthermore, the field
of L2 syntactic processing has been shaped by theories of linguistics
and general psychology which assume an informationally encapsulated
module for language that is separate from other cognitive systems (see
for example Fodor 1983). Most L2 syntactic processing researchers work
under this approach, and they assume – albeit most often tacitly – that
the human sentence processor has a universal architecture which is
subserved by cognitive processes that are specific to language and lan-
guage processing. From this perspective, syntactic processing or parsing
procedures are thought to be universal, and what must be acquired by
child and adult learners alike is the language-specific grammar.3 As noted
above, L2 researchers most often adopt this approach and assume a
universal, modular, serial parser. Other processing models are rarely
taken into consideration by L2 sentence processing researchers, such as
non-modular constraint-based parsers with an interactive architecture,
where all sources of information (linguistic and non-linguistic) are avail-
able in parallel during processing. The constraint-based perspective
assumes that linguistic experience interacts with individual differences
in speakers in, for instance cognitive capacity, and these interactions
might lead to variation in how speakers of different types perform various
linguistic tasks. Therefore as Cook (Chapter 1, this volume) notes, it may
indeed be that constraint-based views are in fact more suitable for the
multi-competence idea of language and language learning. In sum then,
irrespective of whether the universal or constraint-based view of parsing
is assumed, no researcher as yet takes a bilingual perspective as set out
within the multi-competence framework, and no model fits with the
multi-competence perspective that ‘all the languages form part of one
overall system, with complex and shifting relationships between them,
affecting the first language as well as the others’ (Cook, Chapter 1, this
volume, p. 2). This is in fact a challenge to all those working on the nature
of the parser.
In this chapter, I will briefly set out a summary of main research
findings into L2 syntactic processing, providing the theoretical back-
ground to the selected studies, and demonstrating how the monolin-
gual perspective has been pervasive in the field of L2 processing. I argue
that it is possible to begin to move away from the monolingual perspec-
tive by (i) comparing different types of L2 learner (e.g. from different L1
backgrounds, at different stages of development) to each other, rather
than to monolingual speakers; (ii) comparing individual L2 learners’
processing of the target language (TL) to their processing of their
more dominant language(s), including their ‘mother tongue’; and
(iii) investigating longitudinally the same learners’ processing of input
(preferably in all their languages), and how this might relate to their

Syntactic processing 229
linguistic development. Furthermore, (iv) a move away from examining

‘homogeneous’ groups of learners towards a focus on individual varia-
tion is argued to be more suitable to the multi-competence perspective
on L2 syntactic processing. In sum, although studying L2 sentence
processing from a monolingual perspective has given rise to a number
of interesting findings, it may be the time to assess the assumptions
underlying current views of L2 sentence processing. Taking a multi-
competence view is likely to offer a deeper picture of the nature of
the human sentence processing mechanism in general.
11.2 Syntactic processing
11.2.1 Models of parsing

As noted above, the majority of the research into L2 syntactic proces-
sing takes as its starting point what is known about syntactic processing
or parsing from the monolingual sentence processing literature (see
e.g. van Gompel 2013 for a recent overview of current topics and issues).
Despite the fact that monolinguals are rarer than bilinguals, all models
of sentence processing are based on empirical findings of (assumed)
monolinguals.4 Theories contrast in their architecture (interactive
versus modular) and in how and when different types of information
are put to use during processing (parallel versus serial). Interactive
theories of parsing view language (and language processing) as part of
general cognition. All potentially useful pieces of information (seman-
tic, syntactic, pragmatic, discourse, etc.) are applied to the input at the
same time. This set of models are constraint-based (e.g. McRae and
Matsuki 2013) with parsing essentially involving multiple analyses of
the input being run in parallel. That is, a number of different interpreta-
tions of an input string may be generated at the same time, and then
each will receive more or less activation as the parse evolves. In con-
trast to these interactive-parallel models, those advocating serial parsers
(e.g. the Garden Path Model: Frazier 1978) assume that the architecture
of the language processor is modular, or ‘informationally encapsulated’.
In modular-serial parsers, one analysis of the input is constructed at a
time, with syntactic knowledge being applied in the first instance,
followed by lexical–semantic and then real-world/pragmatic knowledge
to assess the well-formedness of the on-going interpretation. This type
of serial, modular, syntax-first parser is most often assumed by those
working on L2 sentence processing, no doubt because the assumptions
fit well with generative, modular views of language and language
competence and the dichotomy between competence and performance
(e.g. Chomsky 1973). It is clear that the idea that multiple languages are
active, to varying degrees, in the mind of an L2 learner when speaking
any of his/her languages is not highly compatible with this perspective.

230 LEAH ROBERTS
11.2.2 Incremental processing

Irrespective of type, all sentence processing models must account for con-
sistent experimental findings, including the fact that comprehenders incre-
mentally parse the input. That is, on encountering a constituent in
the incoming input, listeners/readers immediately attempt to interpret
it on the basis of their analysis so far, and do not wait until they have
all the information before doing so. Incremental processing entails an
extremely rapid application of different types of information in the input
(both bottom-up lexical–semantic and syntactic knowledge and top-down
discourse–pragmatic and world knowledge), although as noted above, dif-
ferent models will predict when different types of information become
available during the parse. As well as incrementally processing the input,
listeners and readers create expectations of material to come, based on their
(often partial) interpretations. The evidence for such mechanisms can be
found in studies in which it is shown that comprehenders often misanalyse
utterances/sentences. This can be seen in how readers temporarily misun-
derstand sentences such as that in (1), a typical ‘garden-path’ sentence.
(1) a. While the band played the song pleased all the customers.
b. While the band played the beer pleased all the customers.
In such temporarily ambiguous sentences, the noun the song is most

likely to be analysed as the direct object of the optionally transitive verb
played. However, once readers meet the main verb pleased in the input, it
becomes clear that the initial direct object analysis was wrong, because (in
a language like English) this verb requires a subject. Comprehenders have
difficulty at this point, arguably because they need to re-analyse the noun
the song as subject of the main verb pleased, rather than as object of the
preceding verb, and this difficulty can be observed by measuring proces-
sing slowdowns in comparison to a sentence in which the noun is unam-
biguously the subject of the next verb phrase. A slowdown occurs because
working memory (WM) capacity is limited, immediate input analysis
is attempted by the parser in order to free up cognitive space, and in
readiness for the next piece of input (e.g. Frazier 1978; Gibson,
Pearlmutter and Canseco-Gonzalez 1996; Gorrell 1995; Townsend and
Bever 2001; Weinberg 1999).
Working memory also plays a role in the processing of long distance
wh-dependencies like (2), where the fronted item or filler (the doctor) must
be kept in memory across more than one clause, until it can be integrated
later on in the input with its licensing constituent (hit) (e.g. Active Filler
Strategy, Frazier and Flores d’Arcais 1989; Phrase-Structure Approach, Gorrell
1995; Minimal Chain Principle, de Vincenzi 1991). The longer the distance
between the fronted item and its licensing constituent – and the greater
the number of intervening constituents – the heavier the load on working
memory (Gibson et al. 1996).

(2) That’s the doctori that the nurse claimed that the angry patient
hit ___i on the ward yesterday.
Evidence for incremental processing in wh-constructions comes in the

form of a ‘filled-gap effect’. For example, in (3b) below, readers slow
down when reading the noun phrase (NP) the pedestrians, when com-
pared to the processing of the same words in (3a). This is because the
verb know can take an animate direct object, unlike declare in (3a) and so
the position immediately following know is the first grammatically
available position for the parser to integrate the fronted constituent.
However, it turns out that this position is filled by an NP already – the
pedestrians – and so processing difficulty is observed in slower reading
times, and the reader must continue reading the sentence until the next
position becomes available, in this case after killed.
(3) a. Who did the police declare the pedestrian killed ___?
b. Who did the police know the pedestrian killed ___?
Interpretation of wh-constructions like (2) and (3) involve establishing a

dependency between constituents, and this process is also necessary in
the interpretation of anaphoric or referring expressions like pronouns
(e.g. it/they). Such constructions are also ambiguous, in that they can refer
to a number of potential referents in the surrounding sentence or dis-
course context or even beyond. Using cues provided by the writer/speaker
(e.g. a pronoun versus a definite NP, he versus the old man) the referring
expression must be linked to the correct antecedent, but there may be
more than one candidate in the context, as illustrated in (4) below.
Processing the pronoun he in (4a) is likely to take more time in comparison
to a condition in which only one of the potential antecedents matches in
number and gender (4b) (Chafe 1976; Dell, McKoon and Ratcliff 1983;
Gordon, Grosz and Gilliom 1993). Also, as is the case with wh-dependency
processing, the process of pronoun resolution involves establishing a link
between a referring expression and its correct antecedent, and this depen-
dency may span multiple clauses or stretches of discourse. This means
that the processor must keep the relevant information active in working
memory at the same time as analysing the intervening input, putting a
strain on memory resources.
(4) a. Johni spoke to Fredj yesterday when hei/j was at work.

b. Johni spoke to Sallyj yesterday when hei/*j was at work.
What the above examples illustrate is that not all sentences are (equally)
easy to understand, even though most of the time we successfully do
comprehend them in real time. Furthermore, different parts of a sentence
might cause different difficulties for (assumed) different reasons. In the
typical L2 syntactic processing study, given that the monolingual perspec-
tive is assumed, how native speakers of the target language process such

232 LEAH ROBERTS
constructions is the benchmark against which L2 learners’ processing per-

formance is measured. Below I set out some key research findings from
the L2 syntactic processing field. I discuss the theoretical underpinnings
of the research and ask whether it is possible to re-envisage such research
within a multi-competence perspective.
11.3 L2 syntactic processing: Theoretical accounts
As noted above, it is uncontroversial that comprehenders incrementally

process the input in real time, making extremely rapid use of informa-
tion from both bottom-up and top-down sources. Since L2 sentence
processing researchers are concerned with comparisons between L2
learners and native speakers, they ask to what extent this is the case
when processing in one’s L2. Most researchers assume a modular
system, with a body of grammatical knowledge and a set of parsing
procedures, and investigate, once grammatical knowledge has been
independently established,5 whether L2 syntactic processing is ‘funda-
mentally’ different from that of native speakers. Clahsen and Felser
(2006) argue for this position and have put forward the Shallow
Structure Hypothesis (SSH). The SSH asserts that L2 learners’ parsing
is lexically driven, particularly during the processing of multi-clause
constructions. Clahsen and Felser argue that unlike native speakers,
who construct detailed, hierarchical syntactic representations during
online sentence processing (e.g. Pritchett 1992), L2 learners use linear
parsing heuristics, relying on lexical–semantic, rather than (abstract)
grammatical information. They do so because ‘the L2 grammar does not
provide the type of syntactic information required to process non-local
grammatical phenomena in native-like ways’ (Clahsen and Felser 2006,
p. 565). In other words, L2 learners can compute local dependencies (for
instance, gender agreement within a noun phrase), but their syntactic
processing is qualitatively different from that of native speakers when
it comes to the parsing of structural dependencies across clauses,
because the L2 grammar is impaired. Other L2 syntactic researchers
argue that L2 learners in fact can syntactically process the input like
native speakers, but it is insufficient L2 proficiency and/or cognitive
capacity limitations, like speed and working memory that cause the
observed processing/performance differences (e.g. Dekydtspotter,
Schwartz and Sprouse 2006; Hopp 2010). This debate parallels that
found in the field of (generative) L2 acquisition as to whether learners
who have acquired their L2 after the Critical Period (Lenneberg 1967)
can ever acquire grammatical knowledge to a native-like level
(Slabakova 2009). The two opposing views are a clear demonstration
of the assumed monolingual perspective in L2 syntactic processing
research as set out in Cook (Chapter 1, this volume): that is, learners

are assumed to be deficient in comparison to native speakers, either in

their grammatical representations (cf. SSH) or their processing capaci-
ties (cf. the quantitative difference account).
Below I briefly discuss some major research findings from the L2 proces-
sing literature to build a picture of what we know to date about how L2
learners parse the input in real time.
11.3.1 Incremental processing and recovery from misanalysis

As noted above, the monolingual approach to L2 processing entails
charting the circumstances under which L2 learners do or do not
parse the input like native speakers. The results of a number of L2
garden-path studies have shown that learners can incrementally pro-
cess the input and that their parsing decisions are often constrained in
similar ways to native speakers. In other words, learners also attempt to
interpret each word of the input as immediately as it is encountered,
and misanalysis can occur, causing processing slowdowns as revision is
required for successful interpretation (Dussias and Cramer Scaltz 2008;
Frenck-Mestre and Pynte 1997; Juffs and Harrington 1995, 1996; Juffs
2004, 2005; Roberts and Felser 2011). An early example comes from
Juffs and Harrington (1995, 1996), who employed a self-paced reading
study combined with a grammaticality judgment task (GJT) with
Chinese learners of English. Like native English speakers, the learners
slowed down when reading the disambiguating verbs (proved/cheated) in
garden-path sentences with pre-posed adjunct (5a) and complement (5b)
clauses, in comparison to non-garden-path sentences which contained
intransitive verbs (e.g. arrive) (5c). This shows that the learners initially
misanalysed the ambiguous NPs (the water, the student) and rapidly
accessed the subcategorization properties of the verbs to constrain
their parsing decisions, i.e. to expect a NP to follow a verb (5a, 5b) or
not (5c).
(5) a. After Bill drank the water proved to be poisoned.

b. Sam warned the student cheated on the exam.
c. After Sam arrived the guests began to eat.
The sensitivity to the subcategorization properties of the verbs during L2

online syntactic processing has been observed in other studies (e.g. Juffs
2004), but where there are L1–L2 differences in argument structure, this
has been found to influence learners’ online processing (Dussias and
Cramer Scaltz 2008; Frenck-Mestre and Pynte 1997).
Therefore, although L2 learners can incrementally process the input,
it is not always clear that they are able to recover the correct inter-
pretation following such misanalysis (Juffs and Harrington 1995, 1996).
One way to assess successful recovery is to look at accuracy scores
on accompanying tasks, and deduce that a low score will suggest

234 LEAH ROBERTS
unsuccessful interpretation of the construction. In other words, low

accuracy can reflect a complete processing breakdown. One can also
use other diagnostics, manipulating the critical construction itself. An
example is the use of semantic information such as ‘pragmatic plausi-
bility’ in garden-path constructions (e.g. Pickering and Traxler 1998).
For example, Roberts and Felser (2011) looked at the word-by-word
reading time profiles of Greek learners of English for sentences such
as (6). The Greek speakers initially attempted to integrate the ambigu-
ous NP (book/girl) with the main verb (wrote). This was demonstrated by
their higher reading times when this initial analysis led to an implau-
sible sentence fragment (wrote the girl) (6b) in comparison to plausible
conditions (wrote the book) (6a). The relative plausibility of this original
fragment affected how easily the learners managed to recover from
their misanalysis. That is, following the disambiguating complement
clause verb (had amazed), the reading time patterns reversed, and pro-
cessing difficulty was observed for the plausible condition (6a). This
shows that the relative (im)plausibility of the sentence fragments (wrote
the book/wrote the girl) affected how committed the learners were to their
initial misanalyses. Like native speakers, commitment to a plausible
initial analysis is stronger than to an implausible one and therefore it is
much harder to abandon in the face of new evidence.
(6) a. The journalist wrote the book had amazed all the judges.
Plausible fragment.
b. The journalist wrote the girl had amazed all the judges.
Implausible fragment.
These learners scored highly in the end-of-trial comprehension questions

in the plausible and implausible conditions (over 90%), which demon-
strates their full recovery from initial misanalyses. However, such recov-
ery was not in evidence in their processing of the more structurally
complex fronted adjuncts (7).
(7) a. While the band played the song pleased all the customers.
Plausible fragment.
b. While the band played the beer pleased all the customers.
Implausible fragment.
Again, the L2 learners were sensitive to the plausibility of the initial

fragments, slowing down at beer in the implausible (7b) versus the plausi-
ble condition song (7a). However, unlike with the complement clause
items (6), the learners’ accuracy scores for the plausible conditions were
significantly lower (85%) than for the implausible condition (95%). The
findings suggest that the plausibility of the original analysis indeed
affected recovery, but at a much later point in processing in comparison
to the complement clause items. The authors argue that the difference in

processing cost arises because preposed adjuncts are structurally (7) more
complex than complement clauses (see Gorrell 1995), which learners find
more difficult to comprehend in real time than native speakers.
Similar processing findings to those in Roberts and Felser (2011) are
observed in L2 processing studies of filler-gap dependencies, like (8)
below. As noted above, given that the comprehender must keep in work-
ing memory the fronted wh-item, while processing the input, so that it is
ready to be integrated with its subcategorizing verb (or at the gap-site),
this type of construction is useful for investigating the potential effects
on sentence processing of individual differences in cognitive capacity.
Dussias and Piñar (2010) investigated the processing of such subject and
object wh-extraction constructions with Chinese L2 learners of English.
Studies of native English speakers show that they initially try to integrate
a filler with the first potential subcategorizer (or from positing a gap at
the first available site), even if the resulting fragment is implausible
when processing wh-dependencies such as Which food did the children read
about __ in class? (e.g. Boland, Tanenhaus, Garnsey and Carlson 1995).
Again, as with garden-path constructions in (6), the plausibility of the
resulting sentence fragment affects how easy it is to recover from this
initial misanalysis (e.g. Clifton 1993; Stowe 1989). The L2 learners in the
Dussias and Piñar study performed in this way, but only those learners
of high working memory, as measured by Waters and Caplan’s (1996)
reading span test.
(8) a. Who did the police know (declare) __ killed the pedestrian?
Subject extraction.
b. Who did the police know (declare) the pedestrian killed __ ?
Object extraction.
The authors found that in the disambiguating region (killed the pedestrian/
the pedestrian killed), all of the learners (and native speakers) found the
subject extraction items overall (8a) more difficult to process at the than
object extractions (8b). This shows that all readers attempted to integrate
the wh-filler into the first clause as object of the main verb, e.g. Who did the
police know? The reading-time slowdowns are arguably caused by having
to reanalyse the wh-filler as subject of the complement clause verb.
However, the learners’ processing patterns differed according to their
scores on the working memory task, with those learners of high working
memory and the native speakers showing an effect of the plausibility of
the filler as direct object. In other words, when the filler was a plausible
direct object (Who did the police know), recovery was more costly than
when it was implausible (Who did the police declare), suggesting that a
plausible initial misanalysis is more difficult to give up than an implausi-
ble one. The learners of lower working memory capacity also attempted
to integrate the filler at the first possible gap-site, as shown by their
much higher reading time for the implausible condition (915 ms) in

236 LEAH ROBERTS
comparison to the plausible condition (730 ms). There was no evidence,

however, that they recovered online from this misanalysis. Similar find-
ings are reported in Jackson and van Hell (2011) for their lower proficient
group of Dutch L2 learners of English and for low working memory
learners in Williams’ (2006) online plausibility judgment study.
The above findings show that there may be different processing profiles
in L2 parsing according to individual differences (IDs) in cognitive capacity
and proficiency. This is potentially interesting if we are to develop a full
picture of L2 parsing; however, there is very little systematic research in
the area of IDs in syntactic processing. This is because most researchers
attempt to minimize potential effects of ID factors rather than make the
study of them central – another illustration of the general theoretical
approach adopted by such researchers. In other words, the concern is
mainly with finding group effects, as is the norm in general psychology.
However, all speakers, at one time or another, find syntactic processing a
strain on cognitive resources for various reasons (talking in a crowded
room, for instance), and it would be beneficial in general to our under-
standing of the nature of the parser to take IDs seriously.
11.3.2 Processing structurally complex constructions

As stated earlier, the major debate in the L2 parsing literature revolves
around the question of whether L2 learners’ syntactic processing will ever
be native-like (Clahsen and Felser 2006). The data from the studies
reported above show that learners are able to use lexical–semantic infor-
mation online, such as plausibility (e.g. Roberts and Felser 2011), animacy
(e.g. Jackson and Roberts 2010), subcategorization (e.g. Juffs and
Harrington 1996) and verb bias information (e.g. Dussias and Cramer
Scaltz 2008). However, learners’ incremental processing, their linking of
a fronted constituent to its lexical licenser, and their ability to successfully
recover from syntactic misanalysis become more difficult the more struc-
turally complex the input is, and at lower levels of proficiency.
As noted above, most L2 parsing researchers assume a modular archi-
tecture for the human sentence processor, making a distinction between
the grammar and the parser. Researchers ask whether learners have access
to and make use of (abstract) syntactic information during online proces-
sing, once they have shown evidence of grammatical knowledge of the
phenomenon. This of course assumes that there is a static body of gram-
matical knowledge and a processing system that has access (or not) to that
knowledge. An example of such a study is Marinis, Roberts, Felser and
Clahsen (2005). The authors used self-paced reading to investigate whether
L2 learners – who had shown evidence in offline tasks of having acquired
the relevant syntactic knowledge – could make use of abstract grammati-
cal information during the processing of long-distance wh-dependencies.
The critical items involved sentences like (9). In (9a), the fronted object the

nurse needs to be linked with the embedded verb angered for comprehen-
sion to be achieved. Critically, for native English speakers, this process
was facilitated by the availability of a purported intermediate syntactic
gap at the clause boundary (ei′). That is, reading times were faster
following angered in (9a), in comparison to object-extraction sentences
that did not contain an intermediate gap (9b) (see also Gibson and
Warren 2004).
(9) a. The nurse [whoi the doctor argued [e′i that the rude patient had
angered ei]] is refusing to work late.
b. The nurse [whoi the doctor’s argument about the rude patient
had angered ei] is refusing to work late.
The authors argue that the L2 learners know English wh-constructions, but
what native speakers – but not L2 learners – can do is activate the fronted
constituent at the grammatically determined position (the intermediate
trace position). This reactivation in memory of the fronted item makes it
more easily available to be integrated with the subcategorizing verb
(angered) in (9a) in comparison to (9b).6
On the basis of these and other studies (e.g. Papadopoulou and Clahsen
2003), Clahsen and Felser proposed the Shallow Structure Hypothesis,
which claims that learners cannot compute abstract grammatical material
during online processing. In the Marinis et al. study, there were no differ-
ences in the processing cost of the two conditions for any of the learner
groups – even those in whose L1 relative clause constructions are formed
in the same way (Greek, German versus Japanese, Chinese). There were
also no effects of differences in proficiency or cognitive capacity (see also
Felser and Roberts 2007). However, those arguing against this fundamental
difference account claim that parallel effects to those of native speakers in
the Marinis et al. study could no doubt be found with sentences like (9) if
the learners had been given enough time (e.g. Dekydtspotter, Schwartz
and Sprouse 2006). This idea has some support if one takes into account
the findings in other L2 parsing studies that show that speed and/or
efficiency differences can be observed in L2 processing in general (e.g.
Fender 2001; Hoover and Dwivedi 1998; Hopp 2006, 2010; Jiang 2004,
2007), and it is in fact the case that the processing of the L2 learners in
the Marinis et al. study (also Felser and Roberts 2007) at later points in the
sentences was not examined.
Debate on L2 grammatical processing also focuses on the extent to
which learners are sensitive to grammatical constraints. An example
comes from Cunnings, Batterham, Felser and Clahsen (2010), who looked
at the processing of islands by German and Chinese L2 learners of English
using eye-tracking during reading. The authors found that the total fixa-
tion times of the native English and the L2 learners showed that neither
group attempted to link a wh-filler inside a relative clause island (10a), only
in the non-island sentences (10b).

238 LEAH ROBERTS
(10) a. Everyone liked the book (city) that the author who wrote con-
tinuously and with exceptionally great skill saw whilst waiting
for a contract. Island condition.
b. Everyone liked the book (city) that the author wrote continu-
ously and with exceptionally great skill about whilst waiting
for a contract. Non-island condition.
In sum, all the L2 learners looked ‘native-like’ in that they respected

relative clause islands, prohibiting the establishment of filler-gap depen-
dencies within them (see also Omaki and Schulz 2011; Juffs 2005; Juffs and
Harrington 1995). These results go against predictions of the SSH, but it is
only assumed that such constraints are in fact grammatical (cf. the
Subjacency Condition: Chomsky 1973). However, one could argue that
such island effects are a reflection of processing difficulties encountered
at the clause boundary of the second relative clause (who wrote). The parser
may slow down here as it attempts to link a filler NP (e.g. the author) with
the relative pronoun while keeping the filler (the book) active in memory
(e.g. Kluender 2004). The results of this study are therefore difficult to
interpret, and demonstrate how it is often problematic to attempt to falsify
the SSH, since many of the proposed structural phenomena that are
investigated could reflect processing rather than grammatical constraints
(see also intermediate trace effects in Marinis et al. 2005).
It is clear that the debate as to whether L2 processing is fundamen-
tally different to that of native speakers sits firmly within the mono-
lingual approach, as Cook describes it. That is, the arguments for and
against the SSH only make sense when learners’ performance is mea-
sured against that of native speakers. One should emphasize that the
speakers in these studies ultimately understand the experimental
items: in fact often they are selected on the basis that they have
demonstrated such ‘offline’ knowledge (see also Roberts and Liszka
2013). One could argue that the monolingual comparison allows for a
structured framework within which to examine L2 processing, and it is
certainly the case that a number of interesting findings have arisen out
of this approach. However, given that the type of phenomena used to
test the hypotheses of the SSH are small in number (e.g. wh-sentence,
reflexive constructions) and often highly complex, examining only the
processing of these can lead to just a partial understanding of (L2)
syntactic processing.
11.3.3 Referential processing

The application of grammatical knowledge comes into play in the processing
of other phenomena, including referential items. An example is Roberts,
Gullberg and Indefrey (2008) who investigated the online interpretation of
Dutch pronouns (11) using eye-tracking and a production task. The results

of the offline production task found that in ambiguous contexts ((11a)

Peter/Hans), the German learners’ offline interpretation for the pronoun
patterned like the native Dutch. They preferred to interpret the pronoun hij
as referring to the most local antecedent (Peter). In contrast, the Turkish
participants’ often resolved the pronoun towards the non-local referent
(Hans). The Turkish leaners were arguably under the influence of their L1:
a null subject language in which an overt pronoun would indicate disjoint,
rather than local co-reference.
(11) a. Peteri en Hans zitten in het kantoor. Terwijl Peteri aan het werk
is, eet hiji een boterham.
(Peteri and Hans are in the office. While Peteri is working, hei is
eating a sandwich.)
b. De Werknemers zitten in het kantoor. Terwijl Peteri aan het

werk is, eet hiji een boterham.
(The workers are in the office. While Peteri is working, hei is

eating a sandwich.)
Therefore, offline, there was an L1 influence effect in the learners’ ultimate

interpretations for the pronouns. In contrast, there was no such L1 effect in
the results of the eye-tracking study. In this case, both the German and
Turkish learners patterned similarly to each other, both slowing down
when reading the subject pronoun (hij ‘he’) in the ambiguous contexts (11a)
in contrast to cases in which there was only one grammatically available
referent for the pronoun (Peter (11b)). Both the learner groups processed these
items differently from the native Dutch speakers, since unlike the Dutch, the
leaners’ processing was disrupted by having an intervening competitor in the
discourse (Hans) (see also Felser, Sato and Bertenshaw 2009). The learners’ L1
can therefore account for the offline but not the online data. The authors
argue that what was different between the learners’ and the native speakers’
processing was that for the learners, processing takes more time, the more
information from multiple sources (e.g. grammatical, discourse–pragmatic)
needs to be coordinated and integrated. This may be the case, irrespective of
how typologically close the learners’ L1 and L2 are (German and Dutch) or the
learners’ final interpretation for the pronoun. In a follow-up study (Roberts,
in preparation), the Turkish bilingual speakers were divided into those who
were dominant in Turkish and those who were classified as Turkish L2
learners of Dutch. Both groups read parallel texts to (11) in Turkish, and
preliminary results patterned together with the Roberts et al. study. That is,
those readers who were less dominant in Turkish were similarly ‘disturbed’
by a discourse competitor as those less dominant in Dutch. These results
suggest that what is often described as native or L1 versus L2 parsing may
more accurately refer to effects of processing in one’s dominant (here, native)
or less dominant language, as is the case for L2 learners.

240 LEAH ROBERTS
11.4 Conclusions: A multi-competence approach to L2

syntactic processing?
The above overview of L2 sentence processing studies demonstrates

that the monolingual perspective is the central approach in the field.
Researchers attempt to identify in what way(s) L2 learners as a group
differ from native speakers, in order to answer theoretical questions
familiar from generative L2 acquisition research. Of course, this involves
assuming that the native speaker processing system is both the norm and
the ideal. How can we move away from a monolingual perspective
towards a multi-competence approach? The starting point may be to
view the multi-competent users as the norm, and to investigate online
processing of different types of user, those who are processing in their less
dominant in comparison to the same speakers processing in their more
dominant language. Processing researchers could look at how processing
is affected at different stages of development, and what the relationship is
between the user’s various languages during real-time processing. The
overview of results in this chapter shows that L2 learners’ performance is
more variable than that of native speakers, and learners are not always
able to access and apply their grammatical knowledge as quickly
(e.g. Hoover and Dwivedi 1998; Jiang 2004, 2007). One could argue, of
course, that given the difference of a few hundred milliseconds, learners’
slower speed of processing is not hugely detrimental, particularly as many
studies show that learners can ultimately interpret the input. However,
of real interest is how these processes develop over time. Longitudinal
studies of L2 processing could specifically chart the progress of learners
as they gain more and more experience.
Another way to approach L2 syntactic processing from a multi-
competence perspective is to rigorously examine individual differences
in L2 parsing, as opposed to attempting to control for them, as is the
norm in most studies. Research shows that when one parses in one’s
less dominant language it puts a burden on cognitive resources, includ-
ing working memory (Dornič 1980), with fewer resources being con-
sumed as L2 proficiency increases (Service, Simola, Metsanheimo and
Maury 2002) and reliable correlations observed between reading span
and phonological memory span scores in the L1 and L2, and with L2
proficiency (Ellis 1996; Harrington and Sawyer 1992; Havik et al. 2009;
Miyake and Friedman 1998; Osaka and Osaka 1992). Taking current L2
parsing research into account, there are mixed findings on the effects of
working memory and L2 syntactic processing (see Juffs and Harrington
2011) no doubt because, as noted above, WM has often been treated as a
variable to be controlled for (e.g. Felser and Roberts 2007; Jackson and
Roberts 2010). Systematically investigating the effect of differences in
cognitive capacity on syntactic processing in one’s more and less

dominant language has the potential to shed light on the role of work-
ing memory in parsing in general, given that in monolingual sentence
comprehension, its role is highly controversial (Baddeley 2003; Traxler,
Williams, Blozis and Morris 2005; Waters and Caplan 1996). Speed is
another potentially influencing factor that is not investigated, even
though it is clear that learners take more time for lexical access, the
accessing of syntactic rules, and applying semantic interpretation to the
language input. As with WM capacity, speed too correlates highly with
proficiency, and reading in one’s less dominant language takes almost
twice the time (e.g. Fender 2001; Kilborn 1992), particularly if reading
in a different script (Marinis et al. 2005; Juffs 2005). However, what is
important is that despite sometimes showing delayed effects (e.g.
Frenck-Mestre 2002, 2005; Hopp 2006; Jackson 2008), being slower
does not seem to affect processing procedures. In fact, perhaps just
because processing is slower in the L2, procedures that are not obser-
vable in ‘monolingual’ studies may become visible (Jackson and Roberts
2010; Roberts and Felser 2011). This is yet another reason to seriously
investigate multilingual speakers, in their various languages.
Another reason to attempt to take a multi-competence approach to L2
parsing is that in the main, most studies focus on what learners know at a
certain point in their development (‘intermediate’, ‘advanced’), and how
their knowledge is used in real-time comprehension. That is, a picture is
drawn of a static moment in time. However, the L2 learners in L2 sentence
processing studies are of course still learning. This raises a serious theore-
tical question that most L2 parsing researchers do not engage with. The
question as to what the relationship is between the development of gram-
matical knowledge and real-time syntactic processing. On the one hand, it
would appear a truism that having to process the input with incomplete
knowledge of the L2 must in some way drive learning. Surely it is syntactic
processing that is being referred to in the acquisition literature on ‘trigger-
ing’ (e.g. Fodor 1983). However, there are few precise accounts of how this
triggering process might work in reality, from either a generative or an
emergentist perspective (see e.g. Kidd 2004 on the processing ‘paradox’ in
L1 acquisition), and the problem becomes even more intractable when one
needs to account for the acquisition of more than one other language.7
Within the generative perspective, Berkes and Flynn (Chapter 10, this
volume) have suggested a reworking of the idea of ‘parameter resetting’.
They argue that as traditionally envisaged, triggering implies a switch
from one (L1) parameter setting to another (L2) as a result of acquisition,
and this is highly problematic, particularly if one wants to take the multi-
competence perspective. Monolingual parsing researchers also do not take
seriously the issue of parsing development, and this has led to a real gap in
our knowledge of the nature of the sentence processing mechanism. Those
working on how (child and adult) learners process the input as they develop
are in a position to drive this critical investigation into parsing.

242 LEAH ROBERTS
Related to the above grammar–parser relationship question, another

problem concerns how production and comprehension processes may
be linked in real time. This again is a topic being debated in the
monolingual processing literature (see e.g. Pickering and Garrod 2013),
and the problem is massively more complex when one takes multi-
lingual language use into account. We know that at least on some levels
all one’s languages are activated even if we are speaking only one
language, with language activation being a fluid and dynamic process
(Grosjean and Li 2013), for example parallel activation (Blumenfeld and
Marian 2013; Costa, Caramazza and Sebastián-Gallés 2000; Marian and
Spivey 2003). In the L1 parsing literature, very little attention is given to
work on learners’ parsing, and in the L2 syntactic processing field, few
researchers attempt to link evidence from the available ‘bilingual’
research findings, perhaps mainly because these tend to focus on lexical
processing and also because they are often connectionist in nature (e.g.
de Bot 2004; Dijkstra and van Heuven 1998; Herdina and Jessner 2002;
see de Groot 2011 for an overview). In sum, L2 parsing as a field can be
argued to focus too intently on theory-internal issues, and therefore
misses out by not communicating with researchers in closely related
fields. Not only can those researchers enhance our knowledge, but L2
parsing findings have the potential to enlighten a number of hugely
important theoretical issues, such as the nature of the relationship
between knowledge and processing, and a multi-competence as
opposed to a monolingual approach may well be the way forward.
Notes
1. I will use syntactic processing and parsing interchangeably to refer to the

application of structural and semantic knowledge to the input during
real-time language comprehension.
2. The majority of researchers working on the topic focus on ‘late’ L2
learners, and therefore I will use this term throughout the paper,
although I recognize that in individual studies, the participants might
better be described as ‘bilinguals’.
3. Many researchers do not explicitly state how they conceive of the
grammar and the parser (and its relationship) and this is somewhat of
a problem, particularly as L2 parsing researchers are becoming inter-
ested in how parsing shapes grammatical knowledge (see Roberts, in
preparation, for some discussion).
4. In the main, researchers do not take detailed background information
on their participants as regards their linguistic knowledge, and it is
assumed that the participants are monolinguals, or at least there is no
discussion of participants’ possible knowledge/expertise of other
languages.

5. The assumption here is that grammar and parsing procedures are

two separate systems. We will return to this point later in the chapter,
in the discussion of other theories of the nature of the grammar–parser
relationship.
6. The authors assume a trace-based account. For other theoretical
accounts of this phenomenon see Gibson and Warren (2004).
7. Recently, there has been more of an interest in the specific role of
syntactic processing in pushing L2 grammatical development (see
e.g. Linguistic Approaches to Bilingualism, issue 3, 2014).
References
Baddeley, A. 2003. ‘Working memory and language: an overview’, Journal of

Communication Disorders 36, 189–208.
Blumenfeld, H. K. and Marian, V. 2013. ‘Cognitive control and parallel
language activation during word recognition in bilinguals’, Journal of
Cognitive Psychology 25 (5), 547–567.
Boland, J. E., Tanenhaus, M. K., Garnsey, S. M. and Carlson, G. N. 1995. ‘Verb
argument structure in parsing and interpretation: evidence from wh-
questions’, Journal of Memory and Language 34, 774–806.
Chafe, W. L. 1976. ‘Givenness, contrastiveness, definiteness, subjects,
topics, and points of view’, in C. N. Li (ed.), Subject and Topic, pp. 25–55.
New York: Academic.
Chomsky, N. 1973. ‘Conditions on transformations’, in S. R. Anderson and
P. Kiparsky (eds), A Festschrift for Morris Halle, pp. 232–286. New York: Holt,
Rinehart and Winston.
Clahsen, H. and Felser, C. 2006. ‘Grammatical processing in language
learners’, Applied Psycholinguistics 27, 3–42.
Clifton, C., Jr 1993. ‘Thematic roles in sentence parsing’, Canadian Journal of
Experimental Psychology 47, 222–246.
Costa, A. Caramazza, A. and Sebastián-Gallés, N. 2000. ‘The cognate
facilitation effect: implications for models of lexical access’, Journal
of Experimental Psychology: Learning, Memory and Cognition 26, 1283–
1296.
Cunnings, I., Batterham, C., Felser, C. and Clahsen, H. 2010. ‘Constraints
on L2 learners’ processing of wh-dependencies: evidence from eye move-
ments’, in B. VanPatten and J. Jegerski (eds), Research in Second Language
Processing and Parsing, pp. 87–110. Amsterdam: John Benjamins.
De Bot, K. 2004. ‘The multilingual lexicon: modeling selection and
control’, International Journal of Multilingualism 1, 17–32.
An Introduction. New York and Hove, East Sussex: Psychology Press.
De Vincenzi, M. 1991. Syntactic Parsing Strategies in Italian. Dordrecht: Kluwer
Academic.

244 LEAH ROBERTS
Dekydtspotter, L., Schwartz, B. and Sprouse, R. A. 2006. ‘The comparative

fallacy in L2 processing research’, in M. G. O’Brien, C. Shea and
J. Archibald (eds), Proceedings from the 8th Generative Approaches to Second
Language Acquisition Conference (GASLA 2006), pp. 33–40. Somerville, MA:
Cascadilla Proceedings Project.
Dell, G. S., McKoon, G. and Ratcliff, R. 1983. ‘The activation of antecedent
information during the processing of anaphoric reference in reading’,
Journal of Verbal Learning and Verbal Behavior 22, 121–132.
Dijkstra, T. and van Heuven, W. J. B. 1998. ‘The BIA-model and bilingual word
recognition’, in J. Grainger and A. M. Jacobs (eds), Localist Connectionist
Approaches to Human Cognition, pp. 189–225. Mahwah, NJ: Erlbaum.
Dornič, S. 1980. ‘Language dominance, spare capacity and perceived effort
in bilinguals’, Ergonomics 23, 369–377.
Dussias, P. E. and Cramer Scaltz, T. R. 2008. ‘Spanish–English L2
speakers’ use of subcategorization bias information in the resolution
of temporary ambiguity during second language reading’, Acta
Psychologica 128, 501–513.
Dussias, P. E. and Piñar, P. 2010. ‘Effects of reading span and plausibility
in the reanalysis of wh-gaps by Chinese–English L2 speakers’, Second
Language Research 26, 443–472.
Ellis, N. C. 1996. ‘Phonological memory, chunking and points of order’,
Studies in Second Language Acquisition 18, 91–126.
Felser, C. and Roberts, L. 2007. ‘Processing wh-dependencies in a second
language: a cross-modal priming study’, Second Language Research 23, 9–36.
Felser, C., Sato, M. and Bertenshaw, N. 2009. ‘The on-line application of
binding Principle A in English as a second language’, Bilingualism:
Fender, M. 2001. ‘A review of L1 and L2/ESL word integration skills and
the nature of L2/ESL word integration development involved in lower-
level text processing’, Language Learning 51, 319–396.
Fodor, J. 1983. The Modularity of Mind. Cambridge, MA: MIT Press.
Frazier, L. 1978. On Comprehending Sentences: Syntactic Parsing Strategies.
Doctoral Dissertation, University of Connecticut.
Frazier, L. and Flores d’Arcais, G. B. 1989. ‘Filler-driven parsing: a study of
gap filling in Dutch’, Journal of Memory and Language 28, 331–344.
Frenck-Mestre, C. 2002. ‘An on-line look at sentence processing in the
second language’, in R. Heredia and J. Altarriba (eds), Bilingual Sentence
Processing, pp. 217–236. Amsterdam: Elsevier.
Frenck-Mestre, C. 2005. ‘Ambiguities and anomalies: what can eye move-
ments and event-related potentials reveal about second language
processing?’, in J. F. Kroll and A. M. B. de Groot (eds), Handbook of
Bilingualism, pp. 268–284. Oxford University Press.
Frenck-Mestre, C. and Pynte, J. 1997. ‘Syntactic ambiguity resolution while
reading in second and native languages’, Quarterly Journal of Experimental
Psychology, section A: Human Experimental Psychology 50A, 119–148.

Gibson, E., Pearlmutter, N., Canseco-Gonzalez, E. and Hickok, G. 1996.

‘Recency preference in the human sentence processing mechanism’,
Cognition 59, 21–39
Gibson, E. and Warren, T. 2004. ‘Reading-time evidence for intermediate
linguistic structure in long-distance dependencies’, Syntax 7, 55–78.
Gordon, P. C., Grosz, B. J. and Gilliom, L. A. 1993. ‘Pronouns, names and the
centering of attention in discourse’, Cognitive Science 17 (3), 311–348.
Gorrell, P. 1995. Syntax and Parsing. Cambridge University Press.
Harrington, M. and Sawyer, M. 1992. ‘L2 working memory and L2 reading
skill’, Studies in Second Language Acquisition 14, 25–38.
Havik, E., Roberts, L., van Hout, R., Schreuder, R. and Haverkort,
M. 2009. ‘Processing subject–object ambiguities in the L2: a self-
paced reading study with German L2 learners of Dutch’, Language
Learning 59, 73–112.
Hoover, M. L. and Dwivedi, V. D. 1998. ‘Syntactic processing in skilled
bilinguals’, Language Learning 48, 1–29.
Hopp, H. 2006. ‘Syntactic features and reanalysis in near-native proces-
sing’, Second Language Research 22, 369–397.
Hopp, H. 2010. ‘Ultimate attainment in L2 inflectional morphology:
performance similarities between non-native and native speakers’,
Lingua 120, 901–931.
Jackson, C. N. 2008. ‘Proficiency level and the interaction of lexical and
morphosyntactic information during L2 sentence processing’, Language
Learning 58, 875–909.
Jackson, C. N. and Roberts, L. 2010. ‘Animacy affects the processing of
subject–object ambiguities in L2 processing: evidence from self-paced
reading with German L2 learners of Dutch’, Applied Psycholinguistics 31,
671–691.
Jackson, C. N. and van Hell, J. G. 2011. ‘The effects of L2 proficiency level on
the processing of wh-questions among Dutch second language speakers
of English‘, International Review of Applied Linguistics 49, 195–219.
Jiang, N. 2004. ‘Morphological insensitivity in second language proces-
sing’, Applied Psycholinguistics 25, 603–634.
Jiang, N. 2007. ‘Selective integration of linguistic knowledge in adult
second language learning’, Language Learning 57, 1–33.
Juffs, A. 2004. ‘Representation, processing, and working memory in a
second language’, Transactions of the Philological Society 102, 199–225.
Juffs, A. 2005. ‘The influence of first language on the processing of wh-
movement in English as a second language’, Second Language Research 21,
121–151.
Juffs, A. and Harrington, M. 1995. ‘Parsing effects in second language
sentence processing: Subject and object asymmetries in wh-extraction’,

246 LEAH ROBERTS
Juffs, A. and Harrington, M. 1996. ‘Garden path sentences and error

data in second language processing research’, Language Learning 46,
286–324.
Juffs, A. and Harrington, M. 2011. ‘Aspects of working memory in second
language learning and teaching’, Language Teaching 44, 137–166.
Kidd, E. 2004. ‘Grammars, parsers, and language acquisition’, Journal of
Child Language 31, 480–483.
Kilborn, K. 1992. ‘On-line integration of grammatical information in a
second language’, in R. Harris (ed.), Cognitive Processing in Bilinguals,
pp. 337–350. Amsterdam: Elsevier.
Kluender, R. 2004. ‘Are subject islands subject to a processing account?’, in
V. Chand, A. Kelleher, A. J. Rodrı́guez and B. Schmeiser (eds), Proceedings
of the 23rd WCCFL. Somerville, MA: Cascadilla Press.
Lenneberg, E. 1967. Biological Foundations of Language. New York: Wiley.
Marian, V. and Spivey, M. 2003. ‘Competing activation in bilingual
language processing: within- and between-language competition’,
Marinis, T., Roberts, L., Felser, C. and Clahsen, H. 2005. ‘Gaps in second
language processing’, Studies in Second Language Acquisition 27, 53–78.
McRae, K. and Matsuki, K. 2013. ‘Constraint-based models of sentence
processing’, in R. van Gompel (ed.), pp. 51–77.
Miyake, A. and Friedman, N. 1998. ‘Individual differences in second lan-
guage proficiency: working memory as language aptitude’, in A. Healy
and L. Bourne (eds), Foreign Language Learning, pp. 339–364. Mahwah, NJ:
Lawrence Erlbaum.
Omaki, A. and Schulz, B. 2011. ‘Filler-gap dependencies and island
constraints in second language sentence processing’, Studies in Second
Language Acquisition 33, 563–588.
Osaka, M. and Osaka, N. 1992. ‘Language-independent working memory
as measured by Japanese and English reading span tests’, Bulletin of
the Psychonomic Society 30, 287–289.
Papadopoulou, D. and Clahsen, H. 2003. ‘Parsing strategies in L1 and L2
sentence processing: a study of relative clause attachment in Greek’,
Pickering, M. J. and Garrod, S. 2013. ‘An integrated theory of language
production and comprehension’, Behavioral and Brain Sciences 36,
329–347.
Pickering, M. J. and Traxler, M. J. 1998. ‘Plausibility and recovery from
garden-paths: an eye-tracking study’, Journal of Experimental Psychology:
Learning, Memory and Cognition 24, 940–961.
Pritchett, B. 1992. Grammatical Competence and Parsing Performance.
University of Chicago Press.
Roberts, L., in preparation. Pronoun resolution in Turkish–Dutch
bilinguals.

Roberts, L. and Felser, C. 2011. ‘Plausibility and recovery from garden

paths in second language sentence processing’, Applied Psycholinguistics
32, 299–331.
Roberts, L., Gullberg, M. and Indefrey, P. 2008. ‘Online pronoun resolu-
tion in L2 discourse: L1 influence and general learner effects’, Studies
in Second Language Acquisition 30, 333–357.
Roberts, L. and Liszka, S. A. 2013 ‘Processing tense/aspect-agreement
violations on-line in the second language: a self-paced reading study
with French and German L2 learners of English’, Second Language
Research 29 (4), 413–439.
Service, E., Simola, M., Metsanheimo, O. and Maury, S. 2002. ‘Bilingual
working memory span is affected by language skill’, European Journal of
Cognitive Psychology 14, 383–408.
Slabakova, R. 2009. ‘The Fundamental Difference Hypothesis twenty years
later’, Studies in Second Language Acquisition 31 (2), 155–173.
Stowe, L. 1989. ‘Thematic structures and sentence comprehension’, in
G. Carlson and M. Tanenhaus (eds), Linguistic Structure in Language
Processing. Dordrecht, The Netherlands: Kluwer.
Townsend, D. J. and Bever, T. G. 2001. Sentence Comprehension. The Integration
of Habits and Rules. Cambridge, MA: MIT Press.
Traxler, M. J., Williams, R. S., Blozis, S. A. and Morris, R. K. 2005. ‘Working
memory, animacy, and verb class in the processing of relative clauses’,
Journal of Memory and Language 53, 204–224.
Van Gompel, R. 2013. Sentence Processing. Hove: Psychology Press.
Waters, G. S. and Caplan, D. 1996. ‘The measurement of verbal working
memory capacity and its relation to reading comprehension’, Quarterly
Journal of Experimental Psychology: Human Experimental Psychology 49, 51–79.
Weinberg, A. 1999. ‘A minimalist theory of human sentence processing’,
in S. Epstein and N. Hornstein (eds), Working Minimalism, pp. 283–315.
Williams, J. 2006. ‘Incremental interpretation in second language
sentence processing’, Bilingualism: Language and Cognition 9, 71–81.

12
Language and cognition
in bilinguals
Annette M. B. de Groot
A couple of years ago Benedetta Bassetti and I accepted an invitation from

the American Journal of Psychology to review each other’s books: Language and
Bilingual Cognition (Cook and Bassetti 2011) and Language and Cognition in
Bilinguals and Multilinguals (De Groot 2011). Given the close resemblance of
the two titles, the book review editor of the journal had naturally assumed
that these volumes cover the same content and thought that a cross-review
of both by the respective authors might be particularly informative for the
journal’s readership, revealing both the established knowledge and the
controversies in their common subject of enquiry. Initially I similarly
assumed a largely overlapping content, but while carrying out my part of
the commitment I soon noticed that the two volumes clearly complement
one another, their titles emphasizing different senses of the component
words. To avoid a similar misunderstanding in the readers of this chapter
and to put them on track right away, I will start out demarcating and
organizing the field of research covered by its title, by explaining a few
of the field’s key terms and sketching the various lines of enquiry it
comprises and the monolingual work it is grounded on. The remainder
of this chapter will centre, from various angles, on lexical concepts,
because these appear to constitute a clear interface between language
and cognition and, as such, provide a link between the field’s diverse
sub-areas of study.
12.1 Demarcating and organizing the field
Although the meanings of language and cognition are usually taken for
granted, like most other words they cover different senses. On the one
hand, language refers to the language system, the structural character-
istics of its components and their inter-relations. On the other hand it
refers to language use, that is, to verbal behaviour in its various forms

Language and cognition in bilinguals 249
(e.g. reading and speaking) and the mental processes and knowledge
structures involved in such behaviour. It is harder to describe the cover-
age of cognition in a few broad strokes. I just searched the first pages of a
dozen textbooks on cognitive psychology or cognitive science for a
common definition of cognition. Most of them start out with a concise
characterization of their subject, such as the statement that cognitive
psychology ‘concerns itself with the science of mental life’ (Kellogg
1995, p. 3) or that it ‘concerns itself with the structure and functions
of the mind’ (Hampson and Morris 1996, p. 1), but in the paragraphs to
follow it soon becomes clear that these crisp descriptions cover numer-
ous areas of study and that it is obscure where cognitive psychology
ends and other areas of psychological science start, or where the demar-
cation between cognition and language lies. Eysenck and Keane (1995,
p. 1) state that cognitive psychology ‘deals with a bewildering diversity
of phenomena . . . thus, it is concerned with topics like perception,
learning, memory, language, emotion, concept formation, and thinking’.
It is particularly noteworthy that language is among this list, indicating
that language is not separate from cognition but one of its ingredients.
This is presumably why a distinction is often made between verbal
cognition (language) and non-verbal cognition, though, as clarified
below, even this distinction is a fuzzy one.
Like the meanings of language and cognition, we usually assume the
meaning of bilingualism to be self-evident, the term referring to an indivi-
dual’s ability (and actual practice) of communicating in two languages and
to the linguistic knowledge base that enables this ability (notice that I use
the term to refer to individual bilingualism, not a community’s bilingual-
ism). At the same time we realize that this description is seriously wanting,
if only because it does not specify whether a particular minimal degree of
expertise in both languages is required for a certain individual to be called
bilingual (instead of an L2 learner). Furthermore, there are multiple routes
towards this ability of using two languages (cf. simultaneous, early sequen-
tial, and late sequential bilingualism, or L2 learning in a natural environ-
ment or at school). The route taken and the degree of expertise in both
languages may impact on the depth of knowledge of the two languages,
the way this knowledge is stored in memory, and the nature of bilinguals’
linguistic expressions. In other words, bilingualism is an umbrella term that
covers much heterogeneity in both language representation and language
use. Still, the available evidence suggests that one phenomenon is true of
all bilinguals, including L2 learners: cross-talk between the two language
subsystems is common, so that using either language is influenced by the
other and the linguistic expressions of bilinguals will typically differ from
the corresponding expressions of monolinguals. This is the main tenet in
much of François Grosjean’s influential work (e.g. Grosjean 1989,
‘Neurolinguists beware! the bilingual is not two monolinguals in one
person’) and, indeed, it is the core of the multi-competence idea as

250 A N N ET T E M . B . D E G R O OT
conceived initially (e.g. Cook 2003), before it was extended to its much
broader current coverage.
The collected research covered by the title of this chapter examines the
effects of bilingualism on verbal and non-verbal cognition. One main line
of study has its roots in cross-language research on linguistic relativity, the
idea that the specific way in which a language system has encoded aspects
of the environment affects the thought of its speakers, with the effect that
speakers of different languages think differently. The bilingual research
that builds on this work addresses the question of how linguistic relativity
plays out in bilinguals, whose two languages may contain different struc-
tural solutions for representing the surrounding world. That is, it deals
with the question of how the opposing features of the two language
systems of bilinguals affect their non-verbal cognition. A second main
research line in the study of bilingual cognition centres on verbal cogni-
tion, specifically, on how mastering two languages affects language use,
the mental processing it involves and the knowledge structures it exploits,
and on how the linguistic expressions of bilinguals compare to those of
monolingual language users. The first of these two research lines consti-
tutes the core of Cook and Bassetti’s (2011) volume mentioned above; the
second is the most central theme in De Groot (2011).
Bassetti and Cook (2011) present some further distinctions that greatly
help in structuring this field of research. One is the contrast between micro-
level and macro-level effects of bilingualism on cognition. Micro-level effects
are the effects caused by specific structural contrasts between a bilingual’s
two languages, for instance, the different ways in which French and
English encode motion, Chinese and English encode time, or Greek and
English encode the colour blue. Other examples of micro-level effects
and their specific effects on cognition will be given in the final sections
of this chapter. Macro-level effects are those due to bilingualism per se,
irrespective of the specific structural contrasts that exist between the two
languages concerned. For instance, using two languages daily may have a
beneficial effect on general ‘executive control’ (an umbrella term for
various mental processes involved in the planning and execution of beha-
viours, such as attending to relevant sources of information, inhibiting
irrelevant information, monitoring behaviour and correcting it when it
goes astray). Bilingualism may be advantageous for executive control
because outputting the selected language at any moment in time, while
preventing intrusions from the other language, requires the involvement
of the mental executive-control system (e.g. Bialystok 2009; but see Paap
and Greenberg 2013); bilinguals’ unilingual language use thus trains this
system. Another macro-level effect of bilingualism might be that knowing
two languages may enhance metalinguistic awareness and, consequently,
be advantageous for Theory of Mind (TOM), the ability to assess what goes on
in the mind of others (e.g. Siegal, Frank, Surian and Hjelmquist 2011). A
third is that it may boost divergent thinking, the ability to quickly generate

many alternative responses to a problem (e.g. Kharkhurin 2007; but see

Hommel, Colzato, Fischer and Christoffels 2011 for a qualification). The
first two of these examples of macro-level effects comprise a further
potentially relevant distinction presented by Bassetti and Cook: the effect
on cognition of using versus knowing two languages (with effects on execu-
tive control and TOM, respectively).
A final relevant contrast concerns two different versions of linguistic-
relativity theory. One version, first advanced by Slobin (e.g. 1987) and
dubbed thinking for speaking by him, concerns the idea that speakers of
different languages think differently from each other when they are actu-
ally using language (e.g. speaking) but not necessarily when they are not
involved in language use. This version of linguistic-relativity theory
appears to be generally accepted for the simple reason that conceptualiza-
tion during language use (a type of thinking) must be shaped to fit onto the
grammatical and lexical structures that are available in the target lan-
guage. The other, challenged, version holds that the thought of speakers
of different languages, or the content of their thought processes, also
differs when they are not using language but are exercising some form of
non-verbal cognition. This second version also comes in two versions,
succinctly defined by Athanasopoulos (2009, p. 83) as the ideas that our
language constrains our thought or that it guides our attention to specific
perceptual aspects of the surrounding world, respectively.
To detect an effect of language structure on thought in general (beyond
thinking for speaking), linguistic-relativity researchers often design
experiments wherein the stimulus materials are completely non-verbal
and no overt language responses are invited from the participants, thus
trying to minimize the involvement of language during task performance
as much as possible. The subjects may, for instance, be asked to memorize
pictured objects or scenes, rate the degree of similarity of pairs of pictures
that depict motion events (Czechowska and Ewert 2011) or colours
(Athanasopoulos 2009), or to select the object out of two that is most
similar to a third object, the three objects differing from each other in the
materials they are made of or in form (Athanasopoulos and Kasai 2008).
Though overt verbal stimulation and overt verbal responses may be
successfully expelled from such tasks, it is quite plausible that many of
them still involve language processing. For instance, tasks that require
memorizing or perceiving familiar objects or colours likely involve cov-
ert naming of the presented objects or colours. As illustrated below, the
covertly generated names may be used to perform the task, deliberately
or unintentionally.
There is another reason for assuming that language is still involved in
tasks that are intended to be non-verbal, namely, that many of the building
blocks of thought, concepts, can be argued to belong to the realm of lan-
guage as much as they belong to the realm of general cognition. Armies of
researchers have tried to elucidate the relation between word meaning,

semantics, on the one hand, and concepts on the other hand, defined by
Murphy (1991, p. 11) as, respectively, ‘the semantic component of words . . .
the component of linguistic elements that gives them significance’ and
‘mental representations of coherent classes of entities. Concepts are our
notions of what kinds of objects and events make up the world.’ Based on
sound reasoning and experimentation, Murphy reaches the conclusion
that, one way or the other, word meanings – components of the mental
lexicon and, thus, of the language system – are built out of concepts. Francis
(2005, p. 251) suggests a specification of this view by first proposing a
conceptual system that is shared across languages and cultures and that is
a ‘possibly hardwired aspect of human cognition’, a ‘universe of possible
ideas or concepts that a human can learn or understand’. She then suggests
that all possible concepts have the potential to be expressed in language, but
that languages, and individual speakers of one and the same language, differ
between them with respect to which ones out of this endless variety of
possible concepts are actually instantiated in language. ‘Semantic represen-
tations may be those concepts that are referred to by particular words or
sentences . . . Word meanings, or semantic representations of words, would
be a particular type of concept’ (p. 252), namely a lexicalized concept (or
lexical concept), a concept that is linked with a word and that is referred to by
this word and that, as such, is part of the language system. It is not unrea-
sonable to assume that lexical concepts constitute a major portion of the
concepts recruited during thinking in general, for the very reason that they
are encoded in language and, therefore, readily available mentally. If true,
performance in so-called non-verbal tasks involves language processing as
well. This idea that lexical concepts are central in both language and
thought in general was the reason to characterize them as the interface
between language and cognition above.
The study of lexical concepts (henceforth used interchangeably with
word meanings and often abbreviated to concepts) is tightly anchored in the
study of language and cognition from a monolingual perspective. One
of the major insights from that work is that word meanings are neither
clear-cut nor stable but vary across contexts, time and individuals (e.g.
Aitchison 1987; Barsalou 1982; Barsalou and Medin 1986; Labov 1973;
Rosch 1975). More recent evidence of the variability of word meaning
derives from the study of bilingualism and L2 learning, which shows that
lexical concepts differ between bilinguals/L2 learners and monolinguals,
and between L2 learners at different stages of L2 development. These
claims are substantiated in the next sections, which deal with different
aspects of the lexical concepts of bilinguals/L2 learners: how they develop
with growth in L2 proficiency and differ from monolinguals’ lexical con-
cepts; how they are connected to the associated words in memory and
what this implies for processing either language. In this discussion, the
terms bilingualism and bilinguals will be used in their neutral sense of
referring to the ability to communicate in two languages and to people

possessing this ability, respectively, irrespective of the degree of expertise

in the languages.
12.2 Forms of bilingualism and determinants of bilingual

lexicosemantic organization
The idea that lexical concepts in bilinguals may differ from the corre-
sponding lexical concepts in monolinguals can be inferred from the col-
lected knowledge in a subfield of bilingualism research that tries to
determine how words connect onto their meanings in bilingual memory.
This subfield of study goes back a long time, at least as far as Weinreich
(1953), who distinguished between three types of word-to-meaning map-
ping, suggesting three forms of bilingualism: ‘coordinate’, ‘compound’
and ‘subordinative’. In coordinate bilingualism a word in one language
and its translation in the other language are connected to separate mean-
ings (in psychological terms: the representation of a word in one language
and the representation of its translation in the other language are connected
to separate meaning representations; henceforth I will often sacrifice preci-
sion for brevity, leaving out the representation of from such phrases). So an
English–French bilingual would have a representation for the English
word dog stored in memory, a second for the French word chien, a third
for the meaning of English dog, and a fourth for the meaning of French
chien (notice that I use word to refer to just the form of a word, not to the
combination of form and meaning). In contrast, in compound bilingualism
both words of a translation pair (e.g. dog and chien) map onto one and the
same meaning (‘dog/chien’). Finally, in subordinative bilingualism words
from a new language that is learned do not map onto meanings directly
but indirectly, via the L1 words.
Compound bilingualism implies the existence of lexical concepts (equa-
ted with word meanings above) that differ from the corresponding lexical
concepts (henceforth often abbreviated to concepts) in monolingual speak-
ers of the two languages for the simple reason that the two terms in a
translation pair seldom have exactly the same meaning. Instead, in addi-
tion to sharing meaning elements, each term of a translation pair is also
associated with language-specific meaning components. To illustrate, this
is what I encountered in a special edition on World War I of today’s news-
paper: ‘If English people talk about war, they mean: mud, bombardments,
trenches, World War I. Dutch people associate war with a winter of starva-
tion, occupation, holocaust: World War II’ (NRC, 26–27 April 2014, trans-
lated from Dutch). Mapping both elements of a translation pair onto one
and the same meaning ignores such language-specific differences in word
meaning (so that, for instance, the common concept contains the lan-
guage-specific meaning components of both languages, or it only contains
the meaning elements that are the same for both languages). The

inevitable effect of such mapping is what I previously called ‘semantically

accented’ word use in both languages (De Groot 2011, 2012, 2014). For
similar reasons, the concepts that subordinative bilinguals use for their L2
differ from the corresponding concepts of native speakers of that language
so that these bilinguals’ L2 (but not their L1) is inevitably semantically
accented.
Ervin and Osgood (1954) and Lambert, Havelka and Crosby (1958) pro-
posed that different forms of bilingualism result from different contexts in
which a person can become bilingual. Specifically, they suggested that
compound bilingualism emerges from ‘fused’ language contexts, as in
foreign language learning in a formal school setting that exploits L1 as a
medium of instruction or when a child grows up in a home where both
languages are spoken interchangeably by both parents. They thought that,
in contrast, coordinate bilingualism emerges from acquisition contexts in
which the two languages are used separately from one another, the one in
one type of context (e.g. at home or in one linguistic community), the
other in another type of context (e.g. at work or in another linguistic
community). Ervin and Osgood (1954) call bilinguals of the coordinate
type ‘true’ bilinguals, perhaps because of the above-mentioned conse-
quence of compound bilingualism that it implies semantically accented
word use. Lambert and his colleagues (1958) provided some empirical
evidence in support of the assumed relation between acquisition context
and form of bilingualism, at the same time offering an important qualifi-
cation of the relation: of the participants who had acquired their two
languages in separate contexts, only those who had learned them in
geographically distinct cultures (‘bicultural bilinguals’), not those who
had acquired them in separate contexts within one and the same geogra-
phical area (‘unicultural bilinguals’), showed a pattern of results that
differed from the pattern observed for participants who acquired their
languages in a fused context. Specifically, in bicultural bilinguals the
meanings of the two terms in translation pairs differed more than in all
other participants. Other authors (e.g. Gekoski 1980) provided some addi-
tional evidence supporting a relation between acquisition context and
type of bilingualism, at the same time noticing that the effects, if they
occur at all, are modest in size and in fact ‘not of sufficient magnitude to
support the usefulness of making a theoretical distinction between com-
pound and coordinate bilingualism’ (Gekoski 1980, p. 444).
A likely reason why the relation between acquisition context and type of
bilingualism is weak is that acquisition context is only one of a larger
number of variables that determines bilingual lexicosemantic organization (as
the representation of words and their meanings and how they are con-
nected is sometimes called). In addition to acquisition context, I previously
identified four such variables on the basis of the empirical evidence exist-
ing at the time (De Groot 1995): level of L2 proficiency, L2 learning strat-
egy, word type and the time delay between current and previous L2 usage.

The joint effect of these variables is that pure subordinative, compound, or

coordinate bilingualism is highly unlikely but that different types of word-
to-meaning mappings coexist within the mental lexicon of one and the
same bilingual; in other words, bilinguals’ lexicosemantic memory is a
‘mixed-representational system’ (De Groot 1993) containing, for instance,
both structures of the compound type and structures of the coordinate
type. Interestingly, this idea had already been suggested by Weinreich
(1953) and Ervin and Osgood (1954), the former stating that ‘It would
appear offhand that a person’s or a group’s bilingualism need not be
entirely of type A or type B, since some signs of the language may be
compounded while others are not’ (p. 10), and the latter suggesting that
‘we would expect speakers of more than one language to distribute
themselves along a continuum from a pure compound system to a pure
coordinate system’ (p. 141).
The word-type effects that are consistently obtained when bilinguals are
asked to process words constitute perhaps the most convincing evidence
for a mixed bilingual lexicosemantic system. The word-type variables that
have been examined most are word concreteness, cognate status and word
frequency. Across a variety of tasks (e.g. within- and between-language
word association, within- and between-language semantic priming and
translation) different response patterns are typically observed for concrete
and abstract words, for words with a cognate translation in the other
language and those with a non-cognate translation (see Van Hell and De
Groot 1998 for some of the evidence and references to relevant studies),
and for frequently and infrequently used words. The observed effects are
often explained in terms of different memory representations for different
types of words. Specifically, it is suggested that translation pairs for con-
crete words and cognates more often share a meaning representation in
bilingual memory (or share a larger part of a ‘distributed’ meaning repre-
sentation in memory; see below) than translation pairs for abstract words
and non-cognates. For example, in an English–Dutch bilingual the transla-
tion pair bottle–fles (concrete) more likely shares a meaning representation
than the translation pair despair–wanhoop (abstract), and the translation
pair elephant–olifant (cognate) more likely shares a meaning representation
than the translation pair squirrel–eekhoorn (non-cognate). In terms of the
compound–coordinate distinction: translation pairs for concrete words
and cognates appear to be represented in compound structures relatively
often whereas those for abstract words and non-cognates more often
seem to be represented in coordinate structures. On the assumption
that lexicosemantic representations develop with increases in L2 use
(see Section 12.3), low-frequency words may relatively often be repre-
sented in subordinative structures and high-frequency words in com-
pound or coordinate structures. Irrespective of the validity of these
assertions, the very fact that different types of words give rise to
different response patterns across a variety of experimental tasks per se

suggests that different types of words exhibit different patterns of word-

to-meaning mapping (in other words, are represented differently) in
bilingual memory.
12.3 The development of bilingual lexicosemantic

representations
The way (the representations of) words are connected to (the representa-
tions of) their meanings is not stable but changes with growth in L2
proficiency. A number of researchers examined this development, a
study by Potter, So, Von Eckardt and Feldman (1984) providing the theore-
tical basis for much of this enquiry. In line with the earlier view that the
linkage patterns may differ between groups of bilinguals, Potter and her
colleagues distinguished between two bilingual lexicosemantic organiza-
tions, assuming that the one applied to L2 learners in an early stage of
learning and the other to proficient bilinguals. The two types of word-to-
meaning mapping they assumed appear to be identical to Weinreich’s
(1953) subordinative and compound organizations, but the names they
gave to these mapping patterns emphasize their consequences for the way
bilinguals process words. The word-association model assumes a direct con-
nection between the L1 and L2 words of a translation pair and between the
L1 word and its meaning. As in Weinreich’s subordinative bilingualism,
the L2 word is not directly connected to meaning. The consequence of this
linkage pattern is that comprehension and production of L2 words come
about indirectly, via the corresponding L1 words. The concept-mediation
model instead assumes that the L1 and L2 words of a translation pair are
both directly connected to a meaning representation that is shared
between the two words and that there is no direct connection between
the latter two (cf. Weinreich’s compound bilingualism). The consequence
of this connection pattern is that comprehension and production take
place in the same, direct, way in both languages.
As mentioned, Potter et al. (1984) hypothesized that the word-associa-
tion and concept-mediation models apply to beginning L2 learners and
proficient bilinguals, respectively, direct connections between the L2
words and meanings gradually developing with increases in L2 experience.
But when they compared two groups of bilinguals with different levels of
L2 proficiency on picture naming in L2 and word translation from L1 to L2
(predicting faster translation than picture naming for the group of begin-
ners but equally long response times on both tasks for the proficient
group; see the original study for details), no difference between the per-
formance of the two groups was obtained, the observed response patterns
suggesting that concept mediation held for both. However, in subsequent
studies wherein the rationale and methodology of Potter et al.’s study were
adopted but L2 learners at an even lower level of L2 fluency than those in

the original study were selected as beginners (Chen and Leung 1989; Kroll
and Curley 1988), performance differences between the groups of begin-
ners and the advanced L2 learners were obtained. In agreement with Potter
et al.’s initial hypothesis, the results observed for the non-fluent and
proficient groups were consistent with the word-association model and
the concept-mediation model, respectively. It thus appears that with
increased L2 experience direct connections develop between L2 words
and their meanings, which they share with L1 words. A candidate learning
mechanism underlying the development of such direct connections is
Hebbian learning, which implies that memory units that are activated
simultaneously get connected (e.g. Li and Farkas 2002; Li and Zhao 2015).
During L2 learning this requirement is fulfilled for the memory represen-
tations of L2 words and their meanings.
Other researchers built on these views by constructing more elaborate
models of bilingual lexicosemantic representation and development. The
plausibly most widely known of these, the Revised Hierarchical Model of
Kroll and her colleagues (e.g. Kroll, Michael, Tokowicz and Dufour 2002;
Kroll and Stewart 1994; see Brysbaert and Duyck 2010 and Kroll, van
Hell, Tokowicz and Green 2010 for recent evaluations of the model),
integrated the word-association and concept-mediation models, assuming
both direct links between the two words of a translation pair and direct
connections from both words to a meaning representation shared by the
two (see Figure 12.1). But instead of the one, bidirectional, connection
between the L1 and L2 words assumed in the word-association model,
the Revised Hierarchical Model postulates two unidirectional ones, the
connection from L2 to L1 being stronger than the one from L1 to L2.
Furthermore, the two connections from the L1 and L2 words to the shared
Word L1 L2
Meaning
Figure 12.1 The Revised Hierarchical Model (adapted from Kroll and Stewart 1994). Solid
and dashed lines represent stronger and weaker connections, respectively

meaning differ in strength, the one from the L2 word being the weaker of
the two.
The differences in strength of the various connections within these
memory structures likely results from unequal use of the two languages,
the assumption being that L1 is the previously and currently more fre-
quently used language. Connection strength determines the speed with
which activation spreads along the links connecting word and meaning, so
that, during the initial stages of L2 development, access of an L2 word from
an L1/L2 meaning (in production) and vice versa (in comprehension) rela-
tively often takes place indirectly, via the L1 word. With growth in L2 use
(and, thus, in L2 word use) and the concomitant gradual strengthening of
the direct connections between the L2 words and the L1/L2 meanings this
entails (e.g. through the mechanism of Hebbian learning described above),
lexical access increasingly often exploits the latter.
The Revised Hierarchical Model emphasizes the strength differences
between the various connections in bilinguals’ lexicosemantic representa-
tions and the effects of these strength differences on the way bilinguals at
different levels of L2 proficiency process words in their L1 and L2. Jiang
(2000) has, instead, proposed a developmental model that, in addition to
paying attention to processing, focuses more on the evolving content of
the memory representations, because ‘representation and processes can-
not be studied independently of each other’ (Jiang 2000, p. 47). He
described L2 development in three stages, in each of which a particular
processing pattern is associated with a specific type of content in the L2
lexical entry. In the first ‘formal’ stage, involving word-association proces-
sing, the L2 lexical entry only stores form information (phonological and/
or orthographic). In the second stage, where processing takes place
through concept mediation, semantic and syntactic information from L1
words is copied in or attached to the corresponding L2 lexical entries. In
the third stage the L2 lexical entry contains formal, semantic, syntactic,
and morphological specifications. In the processing structures Jiang pos-
tulated for this stage, Weinreich’s (1953) coordinate bilingualism can be
recognized. Jiang argued that without extensive contextualized exposure
to the L2, many L2 representations fossilize in the second stage, lacking
morphological specifications and with weak connections between forms
and meanings.
In other developmental models attention specifically centred on the
semantic component of the bilingual lexicosemantic representations.
One of these models is the Shared Distributed Asymmetrical Model by
Dong, Gui and MacWhinney (2005, Figure 12.2; but see also the Sense
Model proposed by Finkbeiner, Forster, Nicol and Nakamura 2004, and
the Modified Hierarchical Model by Pavlenko (2009). In agreement with
other authors (e.g. De Groot 1992; Taylor and Taylor 1990; Van Hell and De
Groot 1998), Dong and her collaborators assumed that the meaning of a
word consists of a collection of more elementary meaning elements

Word L1 L2
Meaning
L1 specific common L2 specific

elements elements elements
Figure 12.2 The Shared Distributed Asymmetrical Model (adapted from Dong et al. 2005).
Solid and dashed lines represent stronger and weaker connections, respectively
(meaning is said to be distributed over a set of conceptual elements; see De

Groot 1992, 406–408, for a discussion about the possible nature of these
conceptual elements). The exact set of meaning elements associated with a
word and the strength of the connections between the word and these
meaning elements changes with increasing L2 proficiency. During an early
stage of L2 learning, the L2 words are connected to both the meaning
elements common to the L2 words and their L1 translations and to L1-
specific meaning elements. In other words, at this stage of learning the L2
words adopt the meanings of the corresponding L1 words. As explained
earlier, L2 word use will consequently be semantically accented. In agree-
ment with the Revised Hierarchical Model, in this stage of L2 learning the
strength of the connections between the L2 words and the meaning ele-
ments are weaker than those between the L1 words and the meaning units.
Advancement in L2 learning involves (i) the strengthening of the links
between the L2 words and the meaning elements common to L1 and L2,
(ii) the weakening of the links between the L2 words and the L1-specific
meanings elements, and (iii) the development of L2-specific meaning ele-
ments and the formation and gradual strengthening of connections
between these and the L2 words (with the effect that L2 word use will
gradually become less semantically accented). Interestingly, the authors
furthermore assumed that, during L2 learning, the emerging L2-specific
meaning elements also become weakly connected to the L1 words, so that
L1 word use may come to exhibit a semantic accent as well (for example,
during comprehension, L2 learners would assign a different meaning to
their L1 words than monolingual L1 users would). Dong et al. obtained
support for this developmental trajectory in a study in which L1 Chinese
speakers at two levels of proficiency in L2 English and monolingual English
and Chinese comparison groups provided semantic-similarity ratings on
word sets consisting of Chinese words or their closest English translations.
I ended the first main section of this chapter highlighting an important
conclusion drawn from earlier monolingual research on the nature of

word meanings – that they vary over contexts, time and individuals – and
claimed that the study of bilingualism substantiates this conclusion. When
positing that claim I had in mind the models of bilingual lexicosemantic
development presented in this section, and especially those that focus on
the meaning components of the lexicosemantic representations. Dong
et al.’s (2005) model (but also the other ones that pay special attention to
the meaning components of these representations but were only men-
tioned in passing) clearly implements the notion that word meanings in
bilinguals are variable over time and differ between individual bilinguals.
But I also had the findings of another type of bilingual studies in mind: those
that concentrate entirely on bilinguals’ lexical concepts (which I equated
with word meanings earlier on in this chapter), ignoring the larger lexico-
semantic structures they are part of. Many of these studies were inspired by
the notion of linguistic relativity and how this plays out in bilingualism.
12.4 Linguistic diversity, linguistic relativity

and lexical concepts
The studies discussed in the previous two sections clearly suggest that the
lexical concepts of bilinguals differ from those of monolinguals and
change with growth in L2 proficiency, but they do not reveal the actual
content of these concepts. A separate research field is primarily aimed at
discovering what information bilinguals’ lexical concepts contain and
how these concepts might differ from the corresponding concepts of
monolinguals and develop in tandem with changes in linguistic experi-
ence and expertise. This area of study relates to, on the one hand, the study
of word-to-concept mapping in bilinguals presented above and with, on
the other hand, research on linguistic relativity: the idea (in tribute to its
originators called the Sapir–Whorf hypothesis) that the language we speak
influences our way of thinking and that, consequently, speakers of differ-
ent languages think differently.
This linguistic-relativity hypothesis builds on linguistic diversity, the
fact that languages differ between them in the linguistic means they use
to represent the environment. This shows from their vocabularies, which
vary greatly in size and the aspects of the world they capture. An illustra-
tion provided by Whorf (but disputed by Martin 1986) is the distinctions
made in Eskimo for snow, where for instance the concepts of snow falling
down, snow that has touched the ground and snow hard like ice are
named differently (Carroll 1956, p. 216). Another illustration concerns
the vocabulary of basic colour terms, which in a seminal study examining
just twenty out of the thousands of languages spoken on the planet was
claimed to already vary between two and eleven (Berlin and Kay 1969; see
Anderson 2010 for reasons why an exact count of the world’s languages
cannot be given). A final example is the rich vocabulary in Pintupi, an

Australian aboriginal language, to distinguish between various types of

holes, with separate words for, for instance, a hole in the ground, a hole in
an object, a hole formed by a rock shelf, a special hole in a spear and a
rabbit burrow (Crystal 1987, p. 15). Linguistic diversity also shows from
morphological variability: across languages, verbs and other word cate-
gories may or may not be marked for gender, number, tense, person,
aspect, for instance, and even for eye-witness testimony (there are lan-
guages, such as Turkish and Iranian, that include verb inflections to
indicate whether the event expressed by a verb was witnessed or heard
about by the speaker, a phenomenon called ‘evidentiality’: Johanson and
Utas 2000).
Linguistic diversity implies that learners of different languages must
attend to different aspects of the world to correctly learn their language.
For example, a learner of Pintupi must find out what it is that distinguishes
something called yarla (a hole in an object) from something called mutara (a
special hole in a spear) whereas a learner of English must find out why they
are both called hole. During this learning process, learners of different
languages develop different inventories of lexical concepts: if one and
the same colour spectrum is lexicalized in, say, six basic colour words in
one language but in eight in another, it follows that the meaning coverage
of (at least some of) the separate colour words differs between these two
languages (i.e. the smaller the number of colour words a language has, the
broader the coverage of each of these). But even if the same set of words
describes a particular conceptual domain in two languages, the coverage
of each single word may differ between languages, because the boundaries
between the lexical concepts may differ between the languages. For
instance, a particular object to hold food may usually be called a dish in
one language but a plate in the other, even if both languages have one word
for dishes and one for plates. Similarly, a particular colour may usually be
called red in one language and brown in the other, even if both languages
have one word for red and one for brown.
There is evidence to suggest that grammatical diversity between lan-
guages also causes particular lexical concepts to differ between speakers of
different languages. For instance, in languages with a grammatical num-
ber system that (from the perspective of native speakers of English) treats
inanimate objects as if they were substances rather than countable entities
(this concerns, for instance, Yucatec and Japanese), the material property
of objects appears to be more prominently represented in the associated
concepts than their shape. The opposite applies to languages like English,
which uses the same grammatical number marking for inanimate objects
and animate beings, treating both as countable (e.g. Lucy and Gaskins
2001). (In the former type of languages, nouns for inanimate objects do
not get a plural marker when referring to more than one of these objects
and the nouns cannot be directly preceded by a numeral to quantify the
objects; instead, like substances in English, they are quantified by means

of a classifier. So instead of three chairs it would be something like three

pieces of chair.) Again, the likely reason for these between-language differ-
ences in the content of concepts is that, to learn the distinctions encoded
in the language, the language learner must attend to specific aspects of
the world (for example, learners of English must direct their attention to
the shapes of objects because the shapes signal how many of these same
objects there are). Once learned, the encoded distinctions will be evoked in
language use time after time and, thereby, strengthened each time. In the
words of the master:
users of markedly different grammars are pointed by their grammars

towards different types of observations and different evaluations of exter-
nally similar acts of observation, and hence are not equivalent as obser-
vers but must arrive at somewhat different views of the world.
(Whorf 1940/1956, p. 221; quoted in Slobin 1987, p. 435)
One further example of how between-language grammatical diversity

may differentially affect the content of lexical concepts concerns gramma-
tical gender, a linguistic construct that designates nouns as masculine,
feminine and neuter, often on the basis of their forms. For instance, in
Italian, in which all nouns are either masculine or feminine, almost all
nouns ending with an -a are feminine and those ending with an -o are
masculine. In many languages a noun’s gender governs the choice and
form of other words with which it occurs. For example, in Italian noun
phrases the forms of adjectives and the choice of articles co-vary with the
noun’s gender (la terra nuova versus il parco nuovo). Languages differ greatly
in the grammatical gender distinctions they make and the degree in which
they use specific word markings to index gender, but the relation between
grammatical gender and biological gender is generally largely arbitrary.
This is perhaps best illustrated by the fact that one and the same object,
and even one and the same living creature, can be named with a feminine
word in one language, a masculine word in a second and a neuter word in a
third.
Even though the relation between grammatical and biological gender is
generally weak, there is evidence to suggest that in languages that distin-
guish between grammatically feminine and masculine words and use gram-
matical-gender marking profusely, the concepts for inanimate objects
contain biological-gender information in agreement with the grammatical
gender of the objects’ names (Boroditsky, Schmidt and Phillips 2003). In
other words, in these languages inanimate objects, evidently sexless, appear
to be assigned biological gender with the effect that, for instance, native
speakers of German may think about bridges as things that are elegant,
whereas native speakers of Spanish may regard bridges to be sturdy (bridge
translates into a feminine word in German and into a masculine one in
Spanish). As pointed out by Boroditsky and her colleagues, the reason why
biological-gender information might come to be stored in the concepts for

inanimate objects, even though the grammatical gender of their names does
not systematically relate to any perceivable feminine or masculine feature
of their referents, is that ‘children learning to speak a language with a
grammatical gender system have no a priori reason to believe that gramma-
tical gender doesn’t indicate a meaningful distinction between types of
objects’ (Boroditsky et al. 2003, p. 64). After all, other grammatical distinc-
tions (e.g. grammatical number) do point at observable distinctions in the
world. Therefore, learners of a grammatically gendered language apply
their common language-learning strategy of searching the perceived objects
for the presence of perceptual features that correlate with the critical
grammatical feature, for instance noticing the elegance of the bridge
when the language to be learned is German but its sturdiness when it is
Spanish.
In short, learners of different languages develop different collections of
lexical concepts and the lexical concepts that are shared between a set of
languages do not cover exactly the same content across these languages
(thus substantiating the earlier claim that true translation equivalents are
rare). On the assumption that lexical concepts are the main vehicles of
thought, these facts suggest that speakers of different languages think
differently, perhaps not only in the non-disputed sense of thinking-for-
speaking (Slobin 1987, 1996), but also in the more dramatic sense that
Whorf had in mind: they develop different views of the world that are also
manifest in non-verbal cognition (see Slobin’s quotation from Whorf above).
12.5 Linguistic relativity and bilingual lexical concepts
Until around 2000 the Sapir–Whorf hypothesis had predominantly been

tested in cross-linguistic studies comparing monolingual speakers of pairs
of languages that differ in how they encode a particular aspect of the world
(e.g. space, time, motion, or colour). Since then bilingualism has been
discovered as a useful test case for linguistic relativity because ‘it is quite
possible that bilinguals are the only ones to experience directly the effects
of linguistic relativity’ (Pavlenko 2005, p. 437). The experience of bicultural
bilinguals that switching to another language feels like switching to
another life (Wierzbicka 1985, in Pavlenko 2005; see also Pavlenko 2006)
indeed suggests that bilinguals’ thinking may depend on the language
spoken at the moment.
Bilingual research into linguistic relativity is as diverse as its monolin-
gual cross-language counterpart (see the rich variety of studies reported in
Cook and Bassetti 2011), but a common goal in a substantial portion of this
work is to discover whether the lexical concepts of bilinguals differ from
the corresponding concepts of monolingual users of their languages. In a
way, an affirmative answer to this question was already implied by the
research on bilingual lexicosemantic organization presented above:

Weinreich’s (1953) subordinative bilingualism (or, in terms of processing,

the word-association model of Potter et al. 1984) implies the use of con-
cepts in L2 that differ from the analogous concepts in monolingual speak-
ers of that language and his compound bilingualism (Potter et al.’s
concept-mediation model) implies the use of concepts that deviate from
those in monolingual speakers of both languages. But the bilingual studies
that build on linguistic-relativity research address this question in a spe-
cific way, looking at the effect of a particular lexical or grammatical
contrast between bilinguals’ two languages on the emerging concepts. In
other words, these studies examine micro-level effects of bilingualism on
cognition, more precisely, on what I am assuming to be the main ingre-
dients of thought: lexical concepts.
The assembled studies in this research area consider a varied set of ways
in which bilingual concepts may differ from monolingual concepts (see
Bassetti and Cook 2011). In addition, they explicate the processes that may
give rise to the various types of bilingual concepts, such as transfer, shift or
convergence (see e.g. Pavlenko 2005 for details). The separate studies often
compare the performance of monolinguals and bilinguals on a verbal or
non-verbal task, using stimulus materials that capture a specific gramma-
tical or lexical contrast between the participants’ two languages (for exam-
ple, a difference in the use of grammatical number, tense, or gender, or in
the way the colour spectrum is lexicalized). A number of them investigated
how task performance develops with increasing levels of L2 proficiency
(e.g. Athanasopoulos 2006; Athanasopoulos and Kasai 2008; Cook et al.
2006; Malt and Sloman 2003) or depends on other variables such as
amount of L2 use or length of stay in the L2 country (Athanasopoulos,
Damjanovic, Krajciova and Sasaki 2011). Furthermore, some studies (e.g.
Bassetti 2007; Boroditsky et al. 2003) were designed to rule out the possi-
bility that the critical effect was due to differential cultural (rather than
linguistic) experience of the bilingual and monolingual participants.
Crucially, the performance differences between monolinguals and bilin-
guals and between various types of bilinguals shown in these tasks often
seem to be attributed to differences in their lexical concepts (seem, because
the authors often use terms that are less specific than lexical concepts,
talking for instance about differences between monolinguals and bilin-
guals in cognition, or in cognitive representation).
On the basis of the assembled evidence, it can be concluded beyond
doubt that speakers of different languages produce different response
patterns on a large variety of tasks, verbal and non-verbal, and that the
observed behavioural differences reflect an influence of the different ways
in which specific aspects of reality are encoded across languages (see e.g.
the studies on colour discrimination by Roberson, Pak and Hanley 2008,
and Winawer et al. 2007). Furthermore, the data unmistakably legitimate
the conclusion that the response patterns obtained for bilinguals generally
differ from those exhibited by monolingual speakers of their languages

(e.g. Ameel, Storms, Malt and Sloman 2005; Athanasopoulos 2009; Bassetti
2007; Malt and Sloman 2003). However, this last conclusion may not
always warrant the more far-reaching one that bilinguals’ concepts differ
from those of monolinguals and that this is the reason why the beha-
vioural patterns differ between them.
The reason for this cautionary note is that, at least in verbal tasks (with
linguistic stimulus materials and/or actual language as output), different
response patterns for bilinguals and monolinguals can also result from the
well-established phenomenon that, when instructed to perform a task in
one language, bilinguals cannot simply mentally switch off (de-activate)
the other language. When, for instance, bilinguals are asked to name
pictured objects in one of their languages, the objects’ names in both
languages are activated (e.g. Costa, Caramazza and Sebastian-Galles 2000;
Starreveld, De Groot, Rossmark and Van Hell 2014). So when a Spanish–
Catalan bilingual is presented with a picture of a table and asked to
produce the corresponding name in Catalan, both the Catalan and the
Spanish words for table (taula and mesa) will be activated in memory. In
their turn, both activated names likely activate their translations in the
other language, that is, they activate each other (cf. the direct connection
between the representations of the two words in a translation pair that is
assumed in some of the models of bilingual lexicosemantic representa-
tion described above) as well as other translations each of them may have
in the other language. The effect of this parallel activation in both lan-
guages is that even if a particular bilingual possesses two perfectly native-
like sets of lexical concepts, the response pattern she exhibits in either
language will differ from the one shown by a monolingual speaker of her
languages. In other words (and as illustrated below), different beha-
vioural patterns are not reliable indicators of differences in concept
representation.
12.5.1 Verbal tasks

This idea that behavioural differences between monolinguals and bilin-
guals in verbal tasks like picture naming do not unequivocally indicate
that their concepts differ, but may be caused by parallel activation of word
representations in their two language subsystems, can already be found in
Ervin’s (1961) seminal study on colour naming in Navajo–English bilin-
guals. She wrote:
If a bilingual is asked to name colours in a particular language, he must

suppress intrusions from the wrong language. If implicit responses
occur in the suppressed language, response probabilities in the overt
language will be altered . . . if an implicit response occurs in the sup-
pressed language, it mediates a response in the overt language.
(Ervin 1961, p. 234)

And perhaps most revealing, she talks about a ‘semantic shift in terms’
(p. 234, my italicization), as if suggesting that bilinguals’ verbal responses,
not necessarily their colour concepts, differ from those of monolinguals.
In sum, contrary to the suggestion conveyed by the title of her article
(‘Semantic shift in bilingualism’), Ervin appears to hold the view that the
different colour-naming patterns she obtained for monolinguals and bilin-
guals reflect response variability due to lexical competition in the bilin-
gual language system, not necessarily that the colour concepts of bilinguals
differ from those of monolingual speakers of their languages (let alone
that the data patterns suggest that bilinguals and monolinguals perceive
colours differently, as if their physiology for colour vision differs).
In De Groot (2014) I attempted to explain the results of a varied set of
studies this way. In all these studies verbal tasks were used. The working
hypothesis in that paper was that phonological, grammatical and semantic
accents (defined earlier as differences between the linguistic expressions of
monolinguals and bilinguals) may have two causes. First, becoming bilin-
gual may result in memory representations for specific linguistic units
(e.g. phones, lexical concepts) that differ qualitatively from the related
representations in monolingual memory (through processes of, for
instance, shift, convergence, or restructuring). To illustrate, the represen-
tation of a particular L2 phone may differ from the corresponding repre-
sentation in a monolingual speaker of that language because during L2
learning it has been assimilated with the representation of a closely simi-
lar but not exactly identical L1 phone (e.g. Flege 1987, 2002). Second,
automatic parallel activation in the two mental language subsystems
may cause accented language even if the critical linguistic units stored
there (e.g. phones, lexical concepts, knowledge structures that enable
parsing) are perfectly native-like. From an examination of the selected
studies with these two options in mind, it appeared that an interpretation
of bilinguals’ accented language in terms of parallel, language-indepen-
dent, activation was a plausible alternative to one in terms of qualitatively
different memory representations of bilinguals and monolinguals, even in
cases where the original authors opted for the latter.
Consider, for instance, the study by Boroditsky et al. (2003) introduced
above, which examined the influence of the grammatical gender of the
names for inanimate objects on the corresponding concept representations.
In one of their experiments highly proficient bilinguals with Spanish or
German as L1 and English as L2 were asked to produce English adjectives to
English names for inanimate objects with a feminine name in German and a
masculine name in Spanish or vice versa. The responses reflected the
grammatical gender of the stimulus words in the participants’ L1. For
example, among the adjectives the Spanish–English bilinguals produced
for the word bridge (masculine in Spanish; feminine in German) were big,
dangerous, strong and sturdy; among those the German–English bilinguals
gave to bridge were elegant, fragile and pretty. As mentioned earlier, this

finding indicates that L1 learners of grammatically gendered languages

assign biological gender to the referents of words and build concepts that
contain biological-gender information in agreement with the associated
words’ grammatical gender. But of particular relevance here is that, at
first sight, it also suggests that, during L2 learning, conceptual transfer
occurs from L1 to L2 and that even the L2 concepts of proficient bilinguals
differ qualitatively from those of native speakers of that language.
But an alternative explanation of the results of Boroditsky et al. (2003)
to consider is that the stimulus words’ translations in L1 mediated the L2
responses. So the stimulus word bridge may have activated its Spanish
translation puente which, in turn, activated the Spanish concept for
puente. Next, the latter’s content was ‘read out’ in complying with the
instruction to produce adjectives. In other words, even if Boroditsky
et al.’s participants possessed perfectly native-like L2-English concepts,
their responses would differ from those of native English speakers per-
forming the same task. It appeared that the findings of a number of other
studies on bilingual concepts, all using (partly) verbal tasks, could simi-
larly be explained, among them one in which the participants were asked
to point out the focal area and range for a set of colour names on a colour
chart (Caskey-Sirmons and Hickerson 1977), a few that examined the
naming of pictured objects (Ameel et al. 2005; Malt and Sloman 2003),
and Ervin’s (1961) study on colour naming in Navajo–English bilinguals
discussed above.
This alternative account of the results of earlier studies presupposes the
existence of direct memory connections between the two word terms in
translation pairs (also in bilinguals of the coordinate type, with a separate
store of concepts for each language) along which activation can be trans-
mitted. Such connections may be formed by Hebbian learning, the learn-
ing mechanism introduced on p. 257. As mentioned there, this mechanism
implies that simultaneously activated memory units get connected and
could thus explain how direct connections between words and their mean-
ings are formed and gradually strengthened. The parallel activation of
both names for a concept during bilingual speech production (as exam-
ined, for instance, by means of picture naming; see above) fulfils this
requirement of simultaneous activation of the two elements in translation
pairs.
12.5.2 Non-verbal tasks

Earlier in this chapter I claimed that so-called non-verbal tasks may still
involve language processing, something researchers using such tasks are
well aware of (e.g. Athanasopoulos et al. 2011, p. 16; Boroditsky et al. 2003,
p. 76; Winawer et al. 2007). Indeed, given the facts that (i) memory con-
nections exist between (the representations of) words and their meanings
and between (the representations of) the two words within translation

pairs, and that (ii) the transmission of activation along these connections
proceeds automatically, tasks that are completely non-verbal but use sti-
muli whose mental representations are linked with word representations
(e.g. pictures of familiar objects or colour patches) involve the activation of
word representations and, thus, language processing. In bilinguals word
representations in both language subsystems are activated in this way (see
above). If activation in a word representation exceeds a critical threshold,
the associated word becomes available to consciousness, that is, internal
verbalization occurs. The observed critical effect might then result from
strategic or unintentional use of the available verbal code(s) rather than
from some fascinating difference in (real) non-verbal cognition in speakers
of different languages or in bilinguals as compared with monolinguals.
Consider for instance a study by Athanasopoulos and colleagues (2011)
that exploited a difference in the way the blue area of the colour spectrum
is lexicalized across languages. More precisely, this study made use of the
fact that some languages like English have just one single word to denote
all shades of blue whereas other languages (e.g. Greek, Japanese and
Russian) have two different single words for blue, one denoting its lighter
shades, the other its darker shades. In each trial two blue colour chips were
shown to the participants, who were asked to judge the degree of similar-
ity of the two on a 10-point scale. The participants were Japanese and
English monolinguals and Japanese–English bilinguals. In some trials the
two stimuli crossed the Japanese boundary between the two categories of
blue (for example, one would be called ao, dark blue in Japanese, the other
mizuiro, a lighter blue). In other trials the two stimuli were instances of the
same category (for example, both would be considered instances of ao).
Crucially, the actual physical distance between the two types of blue
presented in a trial did not differ between these two conditions.
The similarity ratings of the Japanese monolinguals depended on
whether or not the two types of blue presented in a trial crossed the
boundary between the two categories. Specifically, the two types of blue
were regarded as less similar if they belonged to different categories. In
contrast, English monolinguals (for whom the critical category distinction
does not exist) produced equal similarity ratings to within- and across-
category colour pairs. The results of the Japanese–English bilinguals dif-
fered from those of both monolingual groups but resembled those of the
English group most. The results of an earlier Greek–English study showed
similar results (Athanasopoulos 2009).
It is tempting to conclude that perceptual discrimination of colour pairs
depends on whether or not the perceiver’s native language has lexically
encoded the distinction in question. But a more down-to-earth explanation
of the results is that they ensue from automatic verbalization of the colour
stimuli that affects the Japanese and English monolinguals differently.
Specifically, in Japanese speakers two stimuli that cross the boundary
between the lighter and darker shades of blue will internally generate

two different names (ao and mizuiro), whereas two stimuli from the same
category will generate the same name twice (ao or mizuiro). The monolin-
gual Japanese participants may strategically use this information in produ-
cing the requested similarity rating, assigning lower similarity ratings to
stimulus pairs that generate different names than to pairs generating the
same name twice. Because all stimuli give rise to the verbal response blue in
monolingual English speakers, this strategy cannot be sensibly used by
them. The data pattern observed for the Japanese–English bilinguals may
result from cross-trial differences in the moment in time the Japanese and
English colour names (activated in parallel; see above) are available. The
reason why the data resemble the English monolinguals’ data more than
the Japanese data may be that the task instruction was in English (the
participants were asked ‘how different or similar these two colours are’;
Athanasopoulos et al. 2011, p. 13) and were tested in the United Kingdom.
We know that such aspects of an experiment affect the activation level of
bilinguals’ two language subsystems (e.g. Grosjean 1998). The instruction
and test location of the present study have likely boosted the activation in
the English subsystem with the effect that the one English colour term was
more available than the two Japanese terms.
Though it may seem contrived, this interpretation of the data in terms of
the strategic use of internally generated verbal codes receives support
from a similar study that examined within- and cross-category discrimina-
tion of various shades of blue by Russian and English native speakers
(Winawer et al. 2007; no bilingual group was included). Like Japanese,
Russian splits up the blue part of the colour spectrum into two categories,
goluboy and siniy. Instead of the subjective-judgment task used by
Athanasopoulos and colleagues (2011), the researchers used an objective,
perceptual task (speeded colour discrimination with response time mea-
surement), arguing that especially subjective tasks like the one used in
Athanasopoulos’ studies foster the strategic use of verbal codes. But the
presently most interesting manipulation was one that enabled verbaliza-
tion of the colour chips in one condition but frustrated it in a second, the
‘verbal-interference’ condition (in which the participants were asked to
rehearse a series of eight digits while performing the colour-discrimination
task). In the no-interference condition the Russian native speakers, but not
the English native speakers, discriminated between the blue chips faster
when they crossed the category boundary than when they were instances
of the same category, but this effect disappeared in the verbal-interference
condition. This finding suggests the ubiquity of verbalization: it even
occurs when the task used was carefully designed to prevent it, except
when it is frustrated by a secondary task.
But the most important suggestion arising from Winawer et al.’s study
(2007) is that, although speakers of languages that split up the colour
spectrum differently clearly have different stocks of lexical colour concepts,
this specific cross-language difference does not seem to affect perceptual

discrimination of colours. After all, if it did, the Russian native speakers

should also have shown the critical effect in the verbal-interference condi-
tion. Instead, as suggested by Rosch Heider (1972), colour discrimination
may depend on the physiology of primate vision and thus be universal.
12.6 Conclusions
In this chapter I have looked at lexical concepts in the same way Joni
Mitchell (1969) eyed clouds: from multiple angles. I chose to focus on
lexical concepts because they can be argued to belong to the realms of
both language and cognition, both of which I was supposed to cover in this
chapter, from a bilingual perspective. Without a common theme that
assignment would have been utterly impossible. Although the job did
not lead to the kind of illusory observations evoked by Mitchell’s clouds,
like flows of angel hair, ice cream castles and feather canyons, one specific
research area was identified that may be creating an illusion, namely, that
differences in the way different languages lexicalize the colour spectrum
cause their speakers to perceive colours differently. An alternative, more
mundane, interpretation of the critical empirical data was suggested,
namely, that strategic or unintentional use of the output of (apparently
unavoidable) internal verbalization caused the patterns of results to differ
between conditions.
A more general point to make here is that the data pattern obtained in a
particular study may be ambiguous, allowing more than one interpreta-
tion. Indeed, data indeterminacy was also the reason why differences in
the response patterns observed for monolinguals and bilinguals in verbal
tasks like picture naming and colour naming were not regarded as unequi-
vocal evidence that the implicated lexical concepts differed between
monolinguals and bilinguals. Instead, it was suggested that parallel activa-
tion in bilinguals’ two language subsystems and ensuing differences in
response probabilities between bilinguals and monolinguals might be one
cause of the different response patterns observed for them (possible differ-
ences in lexical concepts being a second).
Notwithstanding these cautionary notes, one conclusion can safely be
drawn on the basis of the above discussion: language cannot easily be
banned from non-verbal tasks. In fact, that it cannot was a starting point
in this chapter, which I began by assuming that lexical concepts, elements
of the language system, are a main ingredient of all thought processes. But
the final section of this chapter suggested a stronger presence of language
in non-verbal tasks: the word forms associated with the lexical concepts
may also be recruited during task performance. In fact, this being the case
fully agrees with the structure and processing assumptions contained by
the models of lexicosemantic representation presented in the earlier sec-
tions of this chapter.

References
Aitchison, J. 1987. Words in the Mind: An Introduction to the Mental Lexicon.

Oxford: Basil Blackwell.
Anderson, S. R. 2010. ‘How many languages are there in the world?’
Linguistic Society of America Brochure Series: Frequently Asked Questions.
Washington, DC: Linguistic Society of America.
Athanasopoulos, P. 2006. ‘Effects of the grammatical representation of
9, 89–96.
Athanasopoulos, P. 2009. ‘Cognitive representation of colour in bilinguals:
the case of Greek blues’, Bilingualism: Language and Cognition 12, 83–95.
Japanese blues’, Bilingualism: Language and Cognition 14, 9–17.
Barsalou, L. W. 1982. ‘Context-independent and context-dependent infor-
mation in concepts’, Memory and Cognition 10, 82–93.
Barsalou, L. W. and Medin, D. L. 1986. ‘Concepts: static definitions or context-
dependent representations?’, Cahiers de Psychologie Cognitive 6, 197–202.
Bassetti, B. 2007. ‘Bilingualism and thought: grammatical gender and con-
cepts of objects in Italian–German bilingual children’, International
Journal of Bilingualism 11, 251–273.
Berlin, B. and Kay, P. 1969. Basic Color Terms: Their Universality and Evolution.
Berkeley and Los Angeles: University of California Press.
Bialystok, E. 2009. ‘Bilingualism: the good, the bad, and the indifferent’,
Boroditsky, L., Schmidt, L. A. and Phillips, W. 2003. ‘Sex, syntax, and seman-
tics’, in D. Gentner and S. Goldin-Meadow (eds), Language in Mind: Advances
in the Study of Language and Thought, pp. 61–79. Cambridge, MA: MIT Press.
Brysbaert, M. and Duyck, W. 2010. ‘Is it time to leave behind the Revised
Hierarchical Model of bilingual language processing after fifteen years
of service?’, Bilingualism: Language and Cognition 13, 359–371.
Carroll, J. B. (ed.) 1956. Language, Thought, and Reality: Selected Writings of
Benjamin Lee Whorf. Cambridge, MA: MIT Press / New York and London:
Wiley.
Caskey-Sirmons, L. A. and Hickerson, N. P. 1977. ‘Semantic shift and bilin-
gualism: variation in the color terms of five languages’, Anthropological

Chen, H.-C. and Leung, Y.-S. 1989, ‘Patterns of lexical processing in a

nonnative language’, Journal of Experimental Psychology: Learning, Memory,
and Cognition 15, 316–325.
Cook, V. J. 2003. ‘The changing L1 in the L2 user’s mind’, in V. J. Cook (ed.),
Effects of the Second Language on the First, pp. 1–18. Clevedon, UK:
York and Hove: Psychology Press.
Japanese L2 users of English’, International Journal of Bilingualism 10, 137–152.
Costa, A., Caramazza, A. and Sebastián-Gallés, N. 2000. ‘The cognate facili-
tation effect: implications for models of lexical access’, Journal of
Experimental Psychology: Learning, Memory, and Cognition 26, 1283–1296.
Crystal, D. 1987. The Cambridge Encyclopedia of Language. Cambridge University
Press.
De Groot, A. M. B. 1992. ‘Bilingual lexical representation: a closer look at
conceptual representations’, in R. Frost and L. Katz (eds), Orthography,
Phonology, Morphology and Meaning, pp. 389–412. Amsterdam: Elsevier
Science Publishers.
De Groot, A. M. B. 1993. ‘Word-type effects in bilingual processing tasks:
support for a mixed-representational system’, in R. Schreuder and
B. Weltens (eds), The Bilingual Lexicon, pp. 27–51. Amsterdam and
De Groot, A. M. B. 1995. ‘Determinants of bilingual lexicosemantic organi-
sation’, Computer Assisted Language Learning 8, 151–180.
An Introduction. New York and Hove: Psychology Press.
De Groot, A. M. B. 2012. ‘Bilingualism and cognition’, in C. A. Chapelle (ed.),
The Encyclopedia of Applied Linguistics, pp. 454–462. Oxford: Blackwell.
De Groot, A. M. B. 2014. ‘About phonological, grammatical, and seman-
tic accents in bilinguals’ language use and their cause’, in L. Filipović
and M. Pütz (eds), Multilingual Cognition and Language Use: Processing and
Typological Perspectives, pp. 229–262. Amsterdam and Philadelphia:
John Benjamins.
Dong, Y., Gui, S. and MacWhinney, B. 2005. ‘Shared and separate meanings
in the bilingual mental lexicon’, Bilingualism: Language and Cognition 8,
221–238.
Ervin, S. M. 1961. ‘Semantic shift in bilingualism’, The American Journal of
Psychology 74, 233–241.
Ervin, S. M. and Osgood, C. E. 1954. ‘Second language learning and bilin-
gualism’, in C. E. Osgood and F. Sebeok (eds), Psycholinguistics, supplement
to Journal of Abnormal and Social Psychology, 139–146.

Eysenck, M. W. and Keane, M. T. 1995. Cognitive Psychology: A Student’s

Handbook, third edition. Hillsdale, NJ: Lawrence Erlbaum.
Finkbeiner, M., Forster, K., Nicol, J. and Nakamura, K. 2004. ‘The role of
polysemy in masked semantic and translation priming’, Journal of
Memory and Language 51, 1–22.
Flege, J. E. 2002. ‘Interactions between the native and second language
phonetic systems’, in P. Burmeister, T. Piske and A. Rohde (eds), An
Integrated View of Language Development: Papers in Honor of Henning Wode,
pp. 217–244. Trier: Wissenschaftlicher Verlag.
Francis, W. S. 2005. ‘Bilingual semantic and conceptual representation’, in
J. F. Kroll and A. M. B. de Groot (eds), Handbook of Bilingualism:
Psycholinguistic Approaches, pp. 251–267. Oxford University Press.
Gekoski, W. L. 1980. ‘Language acquisition context and language organiza-
tion in bilinguals’, Journal of Psycholinguistic Research 9, 429–449.
Hampson, P. J. and Morris, P. E. 1996. Understanding Cognition. Oxford:
Blackwell.
Hommel, B., Colzato, L. S., Fischer, R. and Christoffels, I. K. 2011.
‘Bilingualism and creativity: benefits in convergent thinking come
with losses in divergent thinking’, Frontiers in Psychology 2, article 273.
Jiang, N. 2000. ‘Lexical representation and development in a second lan-
guage’, Applied Linguistics 21, 47–77.
Johanson, L. and Utas, B. (eds) 2000. Evidentials: Turkic, Iranian and
Neighbouring Languages. Berlin: Mouton De Gruyter.
Kellogg, R. T. 1995. Cognitive Psychology. Thousand Oaks, CA: Sage Publications.
Kharkhurin, A. V. 2007. ‘The role of cross-linguistic and cross-cultural experi-
ences in bilinguals’ divergent thinking’, in I. Kecskes and L. Albertazzi (eds),
Cognitive Aspects of Bilingualism, pp. 175–210. Dordrecht, Netherlands:
Springer.
Kroll, J. F. and Curley, J. 1988. ‘Lexical memory in novice bilinguals: the role
of concepts in retrieving second language words’, in M. M. Gruneberg,
P. E. Morris and R. N. Sykes (eds), Practical Aspects of Memory: Current Research
and Issues, vol. 2: Clinical and Educational Implications, pp. 389–395.
Chichester, UK: Wiley.
Kroll, J. F., Michael, E., Tokowicz, N. and Dufour, R. 2002. ‘The develop-
ment of lexical fluency in a second language’, Second Language Research
18, 137–171.
Kroll, J. F. and Stewart, E. 1994. ‘Category interference in translation and
picture naming: evidence for asymmetric connections between

bilingual memory representations’, Journal of Memory and Language 33,

149–174.
Kroll, J. F., van Hell, J. G., Tokowicz, N. and Green, D. W. 2010. ‘The Revised
Hierarchical Model: a critical review and assessment’, Bilingualism:
Labov, W. 1973. ‘The boundaries of words and their meanings’, in C.-J. N.
Bailey and R. W. Shuy (eds), New Ways of Analyzing Variation in English,
pp. 340–373. Washington, DC: Georgetown University Press.
Lambert, W. E., Havelka, J. and Crosby, C. 1958. ‘The influence of language-
acquisition contexts on bilingualism’, Journal of Abnormal and Social
Li, P. and Farkas, I. 2002. ‘A self-organizing connectionist model of bilin-
gual processing’, in R. R. Heredia and J. Altarriba (eds), Bilingual Sentence
Processing, pp. 59–85. Amsterdam: Elsevier Science.
Li, P. and Zhao, X. 2015. ‘Computational modeling of bilingual language
acquisition and processing: conceptual and methodological considera-
tions’, in J. W. Schwieter (ed.), The Cambridge Handbook of Bilingual
Processing, pp. 85–107. Cambridge University Press.
Lucy, J. A. and Gaskins, S. 2001. ‘Grammatical categories and the develop-
ment of classification preferences: a comparative approach’, in
M. Bowerman and S. C. Levinson (eds), Language Acquisition and
Conceptual Development, pp. 257–283. Cambridge University Press.
Malt, B. C. and Sloman, S. A. 2003. ‘Linguistic diversity and object naming
by non-native speakers of English’, Bilingualism: Language and Cognition 6,
47–67.
Martin, L. 1986. ‘Eskimo words for snow: a case study in the genesis and
decay of an anthropological example’, American Anthropologist 88, 418–423.
Mitchell, J. 1969. ‘Both sides now’, Beverly Hills, CA: Siquomb Publishing
Company.
Murphy, G. L. 1991. ‘Meaning and concepts’, in P. J. Schwanenflugel (ed.),
The Psychology of Word Meanings, pp. 11–35. Hillsdale, NJ: Lawrence
Erlbaum.
Paap, K. R. and Greenberg, Z. I. 2013. ‘There is no coherent evidence for a
bilingual advantage in executive processing’, Cognitive Psychology 66,
232–258.
Pavlenko, A. 2005. ‘Bilingualism and thought’, in J. F. Kroll and A. M. B. de
Groot (eds), Handbook of Bilingualism: Psycholinguistic Perspectives, pp. 433–
453. Oxford University Press.
Pavlenko, A. 2006. ‘Bilingual selves’, in A. Pavlenko (ed.), Bilingual Minds:
Emotional Experience, Expression, and Representation, pp. 1–33. Clevedon, UK:
Mental Lexicon: Interdisciplinary Approaches, pp. 125–160. Bristol:

Potter, M. C., So, K.-F., Von Eckardt, B. and Feldman, L. B. 1984. ‘Lexical and
conceptual representation in beginning and proficient bilinguals’,
Journal of Verbal Learning and Verbal Behavior 23, 23–38.
Roberson, D., Pak, H. and Hanley, J. R. 2008. ‘Categorical perception of
colour in the left and right visual field is verbally mediated: evidence
from Korean’, Cognition 107, 752–762.
Rosch, E. 1975. ‘Cognitive representations of semantic categories’, Journal
of Experimental Psychology: General 104, 192–233.
Rosch Heider, E. 1972. ‘Universals in color naming and memory’, Journal of
Experimental Psychology 93, 10–20.
Siegal, M., Frank, C. K., Surian, L. and Hjelmquist, E. 2011. ‘Theory of mind
and bilingual cognition’, in V. J. Cook and B. Bassetti (eds), pp. 431–452.
Slobin, D. I. 1987. ‘Thinking for speaking’, Proceedings of the Thirteenth Annual
Meeting of the Berkeley Linguistics Society 13, 435–445.
Slobin, D. I. 1996. ‘From “thought and language” to “thinking for speak-
ing” ’, in J. J. Gumperz and S. C. Levinson (eds), Rethinking Linguistic
Starreveld, P. A., De Groot, A. M. B., Rossmark, B. M. M. and van Hell, J. G.
2014. ‘Parallel language activation during word processing in bilinguals:
evidence from word production in sentence context’, Bilingualism:
Taylor, I. and Taylor, M. M. 1990. Psycholinguistics: Learning and Using
Languages. Englewood-Cliffs, NJ: Prentice Hall.
Van Hell, J. G. and De Groot, A. M. B. 1998. ‘Conceptual representation in
bilingual memory: effects of concreteness and cognate status in word
association’, Bilingualism: Language and Cognition 3, 193–211.
Weinreich, U. 1953. Languages in Contact: Findings and Problems. The Hague,
Netherlands: Mouton.
Whorf, B. L. 1940/1956. ‘Linguistics as an exact science’, Technology Review
43, 61–63, 80–83. Reprinted in J. B. Carroll (ed.), Language, Thought, and
Reality: Selected Writings of Benjamin Lee Whorf, pp. 220–232. Cambridge,
MA: MIT Press.
Winawer, J., Witthoft, N., Frank, M. C., Wu, L., Wade, A. R. and Boroditsky,
L. 2007. ‘Russian blues reveal effects of language on color discrimina-
tion’, Proceedings of the National Academy of Sciences 104, 7780–7785.

13
Gestures in
multi-competence
Amanda Brown
Multi-competence is not restricted to areas traditionally and narrowly

considered to be part of language, but extends to all expressive affor-
dances, including the domain of gesture. Gestures are an integral feature
of communication. As such they potentially become part of what multi-
competent speakers have to acquire in each of the languages used. The
expressive affordances of gestures also mean that they can open a ‘window
onto the mind’, revealing details about multi-competent conceptual repre-
sentations underlying speech.
The increasingly mainstream place of gesture in research on language
and communication is reflected in the growing number of studies exam-
ining gesture in a second language (see overviews in Gullberg 2008, 2012a,
among others). This chapter reviews gesture research conducted in the
fields traditionally known as bi-/multilingualism and second language
acquisition from the perspective of multi-competence. Specifically, we
examine how the frequency, timing, form and semantic composition of
the gestures produced by a multi-competent language user reflect features
of all, some or none of the languages known in dynamic cross-linguistic
interactions and display properties seemingly characteristic of the multi-
competent system.
13.1 An introduction to gesture studies
13.1.1 Definitions and classifications

The contemporary field of gesture studies defines gesture as ‘actions that
have the feature of manifest deliberate expressiveness’ (Kendon 2004,
p. 15), or expressive and communicatively relevant movements related
to ongoing talk. Gestures can be described in terms of their articulatory
and functional properties. Articulatory descriptions label what moves,
how and where (e.g. hand-shape, movement, location in space). They also

Gestures in multi-competence 277
potentially label the anatomy of the movement or ‘gesture phrase’

(Kendon 1980), which comprises optional ‘phases’ such as the ‘prepara-
tion’, ‘retraction’, ‘pre-’ and ‘post-stroke hold’ and the obligatory ‘stroke
phase’, which most see as the gesture itself, in which effortful movement
is concentrated. Functional descriptions, in contrast, characterize the
gesture’s purpose, which is interpreted in the context of the surrounding
speech. Functions of gesture include representing referential content by
iconic, metaphoric or indexical means; organizing interaction by mana-
ging turn-taking or soliciting lexical assistance from an interlocutor; or
indicating rhythmic information. Both articulatory and functional proper-
ties are considered in identifying a gesture as such. A single gesture gen-
erally refers to the most physically effortful and functionally expressive
movement of the hand(s). Additional or alternative articulators may be
involved such as the head or foot; multiple gestures may be produced
simultaneously, for example, by different articulators; and one gesture
may be multifunctional providing, for example, an iconic depiction of an
event superimposed by deictic information.
A number of classificatory systems exist (see Kendon 2004). One of the
most popular (McNeill 1992) employs a series of dimensions, known as
Kendon’s continua, which relate the degree of conventionality in gestural
form to the existence of accompanying speech. Excluding sign languages,
the most conventionalized gestures – emblems – such as the ‘OK’ gesture
produced with the thumb and forefinger linked to form a circle, have clear
standards of well-formedness, are recognizable within speech commu-
nities and can be produced in the absence of speech. Other relatively
conventionalized gestures – deictic or pointing gestures – are generally
accompanied by speech and refer either to concrete referents physically
present in the discourse or to abstract referents not physically present
but assigned a spatial anchor in the gesture space. Less conventiona-
lized gesture types are tightly linked to accompanying speech and
include beats, which are short, pulsating or rhythmic movements used
to reflect the supra-segmental organization of discourse; metaphorics,
which give physical form to abstract concepts such as hand waving to
represent confusion; and iconics, which more transparently depict the
semantic content of speech, such as cupped hands representing a bowl.
These categories are not mutually exclusive and can overlap in a single
movement. The majority of gesture research in multilingualism has
focused on ‘representational’ gestures, i.e. iconics, metaphorics and
deictics (Kita 2000), with some also looking at beats as well as overall
gesture use.
13.1.2 Gestures and speech

Representational gestures are tightly coordinated with speech both
semantically and temporally (Kendon 1972; McNeill 1992; McNeill, Levy

278 AMANDA BROWN
and Pedelty 1990; Schegloff 1984). Regarding semantic coordination, infor-

mation encoded in representational gestures can mirror that encoded in
accompanying speech, highlight certain parts of the spoken utterance, or
even add information not present elsewhere in the discourse. Regarding
temporal coordination, production of representational gestures will typi-
cally be aligned with the most semantically relevant part of the utterance.
To illustrate, while saying the word ‘cup’, a speaker might simultaneously
produce a gesture with a pinched hand-shape and the little finger
extended, denoting the cup as a teacup. With such a close relationship to
speech, representational gestures have been auspiciously labelled as an
additional ‘window onto the mind’ (McNeill 1992), and research suggests
they can offer insights on underlying structural, semantic and conceptual
representations that may be masked by speech. Studies of voluntary
motion, for example, show how gestures may incorporate information
about the left or right direction of movement or the manner in which a
character moves, such as rolling or jumping, which may not have been
explicitly mentioned in the accompanying speech (Kita and Özyürek 2003;
McNeill 2001). Thus, gesture analyses are increasingly part of investigations
of ‘linguistic conceptualization’, where researchers examine the aspects of
an event that speakers choose to select for expression, aspects that tend to
be those readily encoded by their language in speech (or gesture), otherwise
known as ‘event construal’ (von Stutterheim and Nüse 2003), ‘event con-
ceptualization’ (von Stutterheim, Nüse and Murcia-Serra 2002), and ‘think-
ing for speaking’ (Slobin 1996).
An example of the link between speech and representational gesture is
shown in Figure 13.1. Here, a native Japanese speaker describes a scene
from a Sylvester and Tweety cartoon (Canary Row, Freleng 1950), com-
monly used in gesture research, in which Sylvester swings on a rope across
the street. The utterance and the articulatory phases of the gesture phrase
are numbered and transcribed.
In this example, the gesture conveys part of the message encoded in
speech and is deployed at the same time as the relevant part of the
utterance. However, the gesture additionally encodes a right-to-left direc-
tionality, which mirrors the direction of Sylvester’s movement as it was
observed by the speaker (see Kita and Özyürek 2003 for the robustness of
this phenomenon cross-linguistically). In sum, the gesture–speech unit is
conceptually co-expressive.
13.1.3 Gestural diversity across cultures and languages

Gestures exhibit considerable individual variation. Indeed, not everyone
gestures in all contexts. However, systematic differences also exist across
cultures and languages, yielding shared repertoires. The most obvious
culturally based diversity can be found among emblems. Morris et al.’s
(1979) large-scale cross-cultural comparison revealed, for example, that

1biru-kara 2biru-he 3tobi-utsurouto shimasu

building-from building-to fly-tries.to.move do
‘(He) tries to fly from a building to a building’
Figure 13.1 A native Japanese speaker describing a scene from the cartoon Canary Row
(Freleng 1950)
1
Gesture preparation: right hand is raised to a right-side, high-chest position from a resting
position on the lap during production of biru-kara ‘from a building’.
2
Gesture stroke: right hand moves laterally tracing an arc from the speaker’s right to left
during the production of biru-he ‘to a building’.
3
Gesture retraction: right hand is lowered back to a rest position on the lap prior to
production of tobiutsurou-to shimasu ‘tries to fly’.
the same ring gesture, made by connecting the thumb and index finger to
make a circle, indicates acceptance or approval in the US, threatens in
Tunisia, and insults in Brazil. Such diversity is reflected in a number of
cross-cultural inventories (e.g. Johnson, Ekman and Friesen 1981; Morris
1977; Santos 1974; Sparhawk 1981). The articulatory properties of deictic
gestures also vary across cultures, e.g. pointing with the lips versus index
finger (see papers in Kita 2003). Furthermore, cross-cultural differences
have been found in the size of the gesture space, which is larger for
Spaniards than for Germans (Müller 2001). Finally, although the lay belief
that some cultures simply do not gesture has not been supported empiri-
cally, research does suggest that gestural frequency varies by culture (e.g.
Efron 1941/1972; Kendon 2004).
Yet, since gestures are so tightly linked to speech, cross-linguistic differ-
ences in linguistic conceptualization, i.e. the way languages select, encode
and package semantic content in speech, may also be reflected in accom-
panying gestures, yielding gestural variation by language as well as cul-
ture. In investigations of motion, for example, when information about
the trajectory, so-called ‘path’, and manner in which a character moves is
packaged within a clause, as is characteristically the case in English, an

280 AMANDA BROWN
accompanying iconic gesture will likely depict manner and path in a single
conflated movement. However, multiple gestures will generally encode
these features separately if the same information is distributed across
multiple spoken clauses, as is characteristically the case in Turkish (Kita
and Özyürek 2003).
13.1.4 Gestures and acquisition

The importance of the study of gesture has been established for a more
complete understanding of native language use (e.g. Kendon 1986) and of
first language development (e.g. Goldin-Meadow 1998), but Gullberg (2006a)
argues convincingly that a study of gesture is also crucial in research on
second language performance. She distinguishes among the second lan-
guage acquisition of gesture, gestures as a medium of acquisition and
gestures as a window on language acquisition (Gullberg 2012b). In other
words, given that gestural repertoires vary by language and speech commu-
nity, gestures can be viewed as another element of the second language that
learners have to acquire, particularly critical in the case of emblems. In
addition, when interpreted by language users as a source of input, gestures
may facilitate acquisition. Finally, gestures are not produced by everyone or
on every occasion, but when they are, gesture analysis may enable observa-
tion of the way in which second languages are processed, shedding light on
how first and second languages interact within individual minds. This
chapter focuses on the latter, examining what gesture can tell us about
the nature of multi-competence.
13.2 Gestures in multi-competence: Language-specific

patterns
Examinations of the frequency, timing, form and semantic composition of

the gestures produced by multi-competent language users have suggested
the existence of interactions among features of the languages known. Multi-
directional interactions between languages of equal or unequal dominance
are generally considered the norm among bilinguals. However, investiga-
tions of ‘cross-linguistic influence’ or ‘transfer’, defined as ‘the interplay
between earlier and later acquired languages’ (Kellerman and Sharwood
Smith 1986, p. 1), in the field of second language acquisition have largely
focused on one side of this issue – the extent to which features of gestures in
the L2 are shaped by properties of the L1.
13.2.1 Frequency
In counts of gesture strokes as the relevant unit, research has revealed
cultural differences in gesture frequency (e.g. Kendon 2004), and early

ethnographic observations pointed to the implications of such differences

for bi-cultural populations. Efron (1941/1972) described how second-
generation assimilated Southern Italian Americans and Eastern European
Yiddish-speaking Jews in New York city appeared to gesture less frequently
than their first-generation, non-assimilated counterparts. In more recent
quasi-experimental work, when speaking Mandarin, bilingual Mandarin–
English speakers were observed to produce more frequent representa-
tional gestures than monolingual speakers of Mandarin, although
when speaking English, they did not differ from monolingual speakers
of English, who produced more gestures of all kinds, representational
and non-representational, than monolingual speakers of Mandarin (So
2010). These results could be interpreted as arising from a cross-cultural
influence of gesture frequency in English on gesture frequency in Mandarin
within individuals, but problematic in the study is the fact that the Mandarin
speakers were recruited from two different cultural contexts: the monolin-
guals from mainland China and the bilinguals from Singapore, which raises
questions about the source of any cross-cultural or even cross-linguistic
influences since the population of Singapore is considerably multilingual.
Along similar lines, Pika, Nicoladis and Marentette (2006) showed that
French–English and English–Spanish bilingual adults used more iconic ges-
tures when speaking English than did English monolinguals. On the assump-
tion that French and Spanish, like Italian, were higher gesture frequency
languages than English, these results were considered from the perspective
of cross-cultural influence. However, the authors were careful to point out
that neither French nor Spanish monolingual speakers were included in the
study, and that, as a result, cross-cultural transfer could not be separated
from possible general effects of bilingualism (see below). A stronger cultural
effect was observed in Smithson, Nicoladis and Marentette (2011), whose
examination of three groups of children (English monolinguals, Mandarin
Chinese–English bilinguals, and French–English bilinguals) between
seven and ten years old revealed that the French–English bilinguals
gestured significantly more frequently than the Chinese–English bilin-
guals, although in English, monolinguals did not differ from either bilin-
gual group. The researchers concluded that cross-cultural influences may
be a more important determiner of gesture rates among bilinguals than
the general effects of bilingualism and/or working memory capacity.
13.2.2 Timing
Unlike research on gesture frequency, which has examined both of the
languages of functional or advanced-proficiency bilinguals, the majority of
investigations of temporal alignments between gesture and speech have
targeted the L2 of less proficient second language users. These studies have
largely focused on voluntary or caused motion and have generally shown
patterns in the L2 that are reminiscent of the L1. For example, highly

282 AMANDA BROWN
proficient L1 English speakers of L2 Spanish were observed not to align

gestures encoding trajectory, so-called ‘path’ gestures, with verbs as in the
target-like Spanish pattern, but with adverbials, resembling the source-
like English pattern. In contrast, highly proficient L1 Spanish speakers of
L2 English aligned their path gestures with verbs, much like the source
language pattern (Negueruela et al. 2004). Similar findings appeared in
Kellerman and van Hoof (2003) among L1 Spanish learners of L2 English,
although unexpected results for L1 Dutch learners of L2 English (see
below) called into question explanations of the Spanish–English data
based on transfer from the L1.
In an investigation of caused motion, Gullberg (2009a) showed that L1
German speakers of L2 French used the general verbs of placement, e.g.
mettre ‘put’, characteristic of native French discourse, but aligned their
iconic gestures with goal phrases significantly more often than in native
French discourse. At the same time, the L2 speakers aligned significantly
more iconic gestures with verbs than monolingual speakers of German.
Since these patterns were not shared by L1 Dutch speakers of L2 French,
Gullberg argued for an influence of the source language in conceptualiza-
tion of placement, but at the same time a restructuring towards concep-
tualization patterns more typical of native speakers of the L2.
Finally, Stam’s (2006) examination of native Spanish-speaking learners of
English also demonstrated that among advanced speakers, gestures depict-
ing path of motion aligned with verbs significantly more often than in
native English speech. Importantly, in accounting for L2 gesture patterns,
she acknowledged that the data could not tease apart L1 transfer from
development but noted interestingly that lower-proficiency, intermediate-
level learners patterned with native English speakers in this domain. In a
follow-up study nine years later, re-examining the L2 of one of these speak-
ers after extensive immersion experience, alignment of path gestures with
adverbials appeared to have stayed constant, while alignment with verbs
and other elements had moved slightly toward the target L2 pattern (Stam
2010). Moreover, additional analyses of the speaker’s L1 showed a general
similarity to native Spanish discourse, but a possible slight influence of the
L2 reflected in an apparent decrease in the alignment of path gestures with
verbs and noun phrases and an increase in alignment with other speech
elements. Unlike some of the work cited above, this study did not include
the multiple language pairings needed to confirm the existence of cross-
linguistic influence (see Jarvis 2010), but is notable in its inclusion of the L1
and its examination of the development of speech–gesture alignment longi-
tudinally, albeit in an individual case study.
13.2.3 Form
In consideration of cross-linguistic interactions in multi-competent ges-
tural forms, the early study by Efron (1941/1972) provided evidence of less

physically constrained gesture spaces, changes in gesture shapes and

changes in gesture direction from lateral (parallel to or across the body)
to sagittal (perpendicular or away from the body) among assimilated
Yiddish-speaking Jews of Eastern European descent as compared to non-
assimilated Eastern European Jews in New York. Assimilated Italian
Americans were also observed to be more physically constrained and
displayed less head movement than their non-assimilated counterparts.
Although both of the second-generation immigrant groups showed greater
assimilation towards the mainstream American group, they also produced
hybrid gestures, containing properties of both of their cultures simulta-
neously. In a more recent study of gestures depicting motion from L1
Japanese users of L2 English resident in either the L1 or L2 communities,
Brown (2008) revealed patterns of gesture perspective that appeared to
converge between the L1 and L2. Specifically, native speakers of Japanese,
established as functionally monolingual on the basis of self-assessments of
second language proficiency and descriptions of daily language usage,
frequently produced ‘character viewpoint’ gestures, where gestures were
deployed from the perspective of the protagonist, i.e. on a sagittal orienta-
tion with an enactment hand-shape such as the clutching of fingers to
represent the holding of a rope. However, the majority of the multi-
competent gestures in both languages of L1 Japanese users of L2 English
were produced with an ‘observer viewpoint’, where the gesture was
deployed from the observer’s perspective as the action was seen by the
speaker, i.e. on a lateral orientation with a non-enactment hand-shape, a
pattern more prominent in functionally monolingual English discourse.
13.2.4 Semantic composition

With respect to investigations of the semantic content of gestures, most
research again focuses on the domain of motion. In their study of volun-
tary motion, Negueruela et al. (2004) found so-called ‘manner fog’ gestures
(McNeill 2001), which refer to cases in which manner of motion is not
explicitly expressed in speech, but instead included in accompanying
iconic gestures, among highly proficient L1 Spanish speakers of L2
English, much like the pattern seen in native Spanish discourse (see
similar results for Korean–English bilinguals in Choi and Lantolf 2008).
While this suggested an influence of the L1 on the L2, manner fog gestures
were also seen in the L2 production of native English learners of Spanish
(although note different results in the L2 English of Korean L1 speakers in
Choi and Lantolf 2008). Given that such gestures are not characteristic of
the source language, English, on the surface the results for L2 Spanish do
not seem to be attributable to L1 influence. However, one possibility is that
L2 Spanish speakers look for equivalents of the very specialized manner
terms abundant in their L1, English, do not find them in Spanish, and turn
to gesture instead, which yields seemingly target-like performance.

284 AMANDA BROWN
Ultimately, like much of the research above, this study could not distin-
guish the effects of cross-linguistic influence from developmental markers
of lexical acquisition of manner terms.
Other research with L1 Dutch users of L2 Japanese found persistent
inclusion of information about ‘ground’, that is the terrain on which
motion occurs (Talmy 2000), while speaking and gesturing about move-
ment in the L2. This pattern was more typical of the source language,
Dutch, as opposed to the series of statements and accompanying gestures
about ground prior to mention of movement more typical of the target
language, Japanese (Yoshioka and Kellerman 2006).
Similar to the L1 German speakers of L2 French included in her study,
Gullberg (2009a) found target-like speech in L2 descriptions of caused
motion by L1 Dutch speakers of L2 French. However, their gestures in
the L2 encoded spatial properties of the figure object, specifically whether
the shape of the object prompted placement on a horizontal or vertical
orientation, significantly more often than native speakers of French,
though less often than monolingual speakers of Dutch, who routinely
produced such gestures. Given the absence of such patterns among the
L1 German speakers of L2 French, Gullberg again suggested an influence of
the source language in conceptualization of placement, an example of
‘conceptual transfer’ (Odlin 2008) from the L1, but also of a restructuring
towards conceptualization patterns more typical of native speakers of the
L2. Interestingly, although all the L1 Dutch speakers of L2 French had
comparable formal proficiency, the ostensible ‘in-between’ source and
target language performance was more prevalent among the slightly less
fluent speakers, as identified by use of markers of disfluency, specifically
filled pauses. Gullberg discussed how gesture analyses in this context
revealed the challenges involved in restructuring ‘thinking-for-speaking’
(Slobin 1996) in an L2 even when the lexical items and syntactic structures
of the L2 were relatively uncomplicated.
Very little research on cross-linguistic interactions in the semantic com-
position of multi-competent gestures has included simultaneous observa-
tion of the L1. In further analyses of placement events, Gullberg (2009b)
showed that L1 English learners of L2 Dutch maintained L1 patterns in
both L1 and L2 speech with use of neutral constructions depicting trajec-
tories as opposed to those incorporating properties of the figure object
typical of native Dutch. The gestures accompanying those constructions
also depicted path alone. Brown and Gullberg (2008) and Brown (2015), in
contrast, focused on the distribution of information between speech and
gesture, specifically the encoding of information about manner of motion.
The first of these studies showed converging patterns between the L1 and
L2 among L1 Japanese speakers of L2 English at an intermediate level of
proficiency, with varying residence in the target or source language com-
munities. Specifically, regardless of residence, L2 users encoded manner in
speech but only path in accompanying iconic gestures significantly more

often than functionally monolingual speakers of Japanese, though less

often than functionally monolingual speakers of English, a phenomenon
labelled ‘manner modulation’ (McNeill 2001). In the follow-up study, L1
Mandarin speakers of L2 English without residence in the target-language
community and at a comparable level of proficiency to the Japanese–
English speakers also demonstrated converging patterns between the L1
and L2 in so-called ‘manner highlighting’ gestures, where manner was
double-marked in speech and gesture significantly more often in the L1
and L2 by L2 users than by monolingual speakers of Mandarin, but less
than by monolingual speakers of English. To the extent that the patterns
displayed by the L2 users in each of the two groups (Japanese–English and
Mandarin–English) and in each of their L1s and L2 were between the
monolingual patterns in the respective source and target languages and
yet distinct from one another, these results were also interpreted from the
perspective of cross-linguistic influence. In her longitudinal case study of
an L1 Spanish speaker of L2 English, Stam (2010) claimed that gesture
patterns regarding manner of motion remained source-like, with manner
gestures occurring with and without manner in speech in both the L1 and
L2, a pattern not attested in native English, but present in native Spanish
discourse. She also noted changes towards the target language in the
extent to which gestures conflated manner and path. Interestingly, how-
ever, this particular finding was largely ignored in the conclusion, which
instead stressed a seeming lack of development in the learner’s expression
of manner and the persistence of non-target-like thinking-for-speaking in
the L2. Finally, in contrast to the above, Choi and Lantolf’s (2008) case
study of two L1 Korean–L2 English and two L1 English–L2 Korean speakers
found no distinction between bilingual and monolingual L1 English or L1
Korean with respect to the semantic composition of gestures.
13.2.5 Discussion of language-specific patterns from the

perspective of multi-competence
With some exceptions, the majority of studies examining gesture from the
perspective of cross-cultural and cross-linguistic influence have focused on
the L2 only, with evidence of features of gesture frequency, timing, form
and semantic composition purportedly traceable to the L1, although often
without the triangulating data from additional language pairings that
would be needed to fully support such an account (see Jarvis 2010 for
discussion of methodology in this research domain). Gestures in production
are not typically judged as grammatical or ungrammatical (with the excep-
tion of violations in the structural well-formedness of emblem gestures); yet
descriptions of L2 gesture production, particularly in observations of those
described as second language ‘learners’ as opposed to bilinguals, frequently
refer to ‘non-target-like’ or ‘source-like’ behaviour, a ‘lack of acquisition’ or
the ‘maintenance of L1 patterns of thinking-for-speaking/gesturing in the

286 AMANDA BROWN
L2’, which implies the existence of negative transfer from the L1. As dis-
cussed in Chapter 1 of this volume, such descriptions arguably adopt a
‘monolingual perspective’ and reflect a ‘deficit’ view of multilingualism
and second language performance (see e.g. Birdsong 2005), where L2 pat-
terns are compared to those from native speakers, frequently monolinguals,
and found wanting. In addition, many of these same studies only consider
the multi-competent speaker’s L2 and thus are inherently biased, skewed
towards only half of the multi-competent profile, as Cook states (Chapter 1),
‘yin without yang’.
While comparisons between monolingual and L2 performance are cer-
tainly valuable, those that consider all the languages of the multi-competent
speaker, particularly in investigations of cross-linguistic influence, are more
in line with the notion of the multi-competent system. A few of the studies
cited above have concurrently examined the L1 (e.g. Brown 2008; Brown and
Gullberg 2008; Gullberg 2009b; Stam 2010), though such analyses are more
frequently part of observations of functional bilinguals (e.g. Choi and
Lantolf 2008; Pika, Nicoladis and Marentette 2006; Smithson, Nicoladis
and Marentette 2011; So 2010). These studies have generally corroborated
previously described influences of the L1 on the L2 in gesture, but in some
cases revealed the bi-directionality of interactions between the languages
and cultures of a multi-competent speaker. Such findings suggest that the
combination of languages and cultures in multi-competence affects and
conceivably shapes the frequency, form, semantic composition and poten-
tially the timing of gesture production in the L1 as well as the L2, and may
yield a converging or even merged gestural system. The handful of studies
producing this type of evidence support Premise 1 of multi-competence –
the concept of ‘mutual interdependence’ and inter-relationships between
the L1 and L2, as well as the notion of a single gestural system, with proper-
ties of each of the languages/cultures known.
Plenty of avenues for additional work remain. First, as noted through-
out, much of the research needs to be replicated with additional language
pairings in order to distinguish the seemingly bi-directional, cross-
linguistic interactions from more general effects of multi-competence,
discussed below, especially as some of the language-specific effects of
multi-competence are not consistent across studies (e.g. Negueruela et al.
2004 versus Choi and Lantolf 2008). Furthermore, gesture research has
yet to address cross-linguistic interactions in languages beyond the L1
and L2, and the multi-competence framework clearly mandates
consideration of an L3, L4 etc.; see for example de Angelis (2007) and
Gabrys-Barker (2012) for recent discussions of multilingual language
acquisition. Finally, the effects of language proficiency and cultural immer-
sion are not entirely understood. Nicoladis (2007) argues that, like many
aspects of transfer, cross-linguistic influences in gesture patterns might be
more prevalent at lower levels of L2 proficiency as a strategy to enhance the
effectiveness of communication. Indeed, Stam’s (2010) case study described

fewer seemingly cross-linguistic interactions after a lengthy immersion in

the L2 community. Similarly, stronger patterns of L1 influence on the L2
were found among L2 users who, despite sharing comparable L2 proficiency
formally with their peers, seemed slightly less fluent in their L2 (Gullberg
2009a). Furthermore, several studies have revealed a multi-competent ges-
ture profile with potential bi-directional interactions between the source
and target language systems at intermediate levels of formal L2 proficiency
regardless of the level of cultural exposure (Brown 2008, 2015; Brown and
Gullberg 2008), and there is less evidence of cross-linguistic influences at the
level of advanced functional bilingualism (e.g. Choi and Lantolf 2008). These
particular findings imply that, similar to cross-linguistic influences in other
domains such as foreign accent, multi-competent gesture patterns are char-
acterized by stronger cross-linguistic interactions at lower levels of profi-
ciency, and that later development might yield greater separation. Counter
to this prediction, however, is the cross-sectional research demonstrating
evidence of more cross-linguistic interactions at higher levels of proficiency
than at intermediate levels (Stam 2006). Since the multi-competence frame-
work allows for either greater separation or greater integration of language
features over time, further work is needed to observe gesture patterns
longitudinally within and across languages in the development of multi-
competence to assess the impact of proficiency and cultural immersion.
13.3 Gestures in multi-competence: Language-general

patterns
In addition to the potential for cross-linguistic interactions, the frequency,

timing and semantic composition of the gestures produced by a multi-
competent language user may display properties apparently unique to the
multi-competent system. Commonalities in performance features across
different L2s have traditionally been labelled as the ‘learner variety’, but
there has been no comparable term or way to view similarities across the
L1s of multi-competent speakers. Examples of such similarities are con-
sidered next.
13.3.1 Frequency
Conflicting findings exist with respect to how frequently gestures are
produced by multi-competent populations. In a study of French–English
bilingual children, Nicoladis, Pika and Marentette (2009) found that ges-
tures were more frequent among bilinguals in both of their languages than
among monolinguals of each language, arguing that gesture use may
facilitate more demanding lexical retrieval in bilingualism. This study
challenged the assumption made in Pika, Nicoladis and Marentette
(2006) of French being a higher gesture frequency language than English

288 AMANDA BROWN
since the later study showed no differences in gesture rates between

monolingual French and English speakers, at least for children. In addi-
tion, findings from Nicoladis et al. (2009) call into question the interpreta-
tion of Pika et al. (2006), which had attributed the higher rates of iconic
gesture among French–English and English–Spanish bilingual adults as
compared to English monolinguals to cross-cultural influences. Instead,
the later findings suggest that more frequent gestures among bilinguals
may be a more general artifact of bilingualism, not specific to the interac-
tions of particular cultures.
Several studies have shown differential gesture use in bilingualism with
respect to the frequencies of specific gesture types across the L1 and L2. In
an early study examining the use of gestures as a communication strategy
among intermediate-level L2 users, Gullberg’s (1998) analysis of L1
Swedish learners of L2 French and L1 French learners of L2 Swedish
revealed that L2 users generally did not replace lexical units with mimetic
gestures in their L2, but instead produced more iconic gestures in their
dominant languages. On the other hand, L2 speakers produced more
abstract gestures (metaphorics and deictics) in their weaker languages,
which she argued may serve pragmatic functions and help with the gram-
matical organization of discourse. A follow-up study focused on reference
tracking demonstrated that L2 discourse was characterized by over-
specification of maintained referents both in speech and in gesture
(Gullberg 2003). In other words, L2 Swedish and French speakers produced
full lexical NPs in contexts where native speakers would use pronouns or
zero marking, and spatially anchored those referents with abstract deictic
gestures. Results were replicated for L1 Dutch learners of L2 French at
higher levels of proficiency in Gullberg (2006b), where it was also shown
that manipulating visibility between speaker and hearer had no impact on
the extent of over-specification of referents in L2 discourse. Comparable
results have been found with advanced-proficiency L2 speakers across
different languages. Sherman and Nicoladis’ (2004) observation of L1
English users of L2 Spanish and L1 Spanish users of L2 English also
demonstrated more deictic as opposed to iconic gestures in the L2 (see
also Nicoladis 2007 for an overview of studies showing more frequent
iconic gestures in the dominant language of French–English bilingual
preschoolers). Nicoladis et al. (2007) also described more deictic gestures
in the L2 English of Chinese–English bilingual adults than in their L1,
but, in contrast to the above, no difference in iconic gesture use by
language. Furthermore, a closer inspection of the data revealed that
deictic gestures did not appear to be associated with grammatical issues
and conversely that iconic gestures were employed more frequently in
cases of ‘language difficulty’ in the L2. Nagpal, Nicoladis and Marentette
(2011) found a higher gesture rate in intermediate to advanced L2 English
than L1 Hindi, though also no difference between the languages in fre-
quency of iconic gestures. Furthermore, Stam’s (2010) longitudinal case

study documented gestures in L2 English that became less segmented over

time. Whereas earlier in development, ‘almost every grammatical constitu-
ent had its own gesture’ (p. 78), later in development, gestures spanned
more constituents. Finally, in a study not specifically designed to investigate
gesture frequency, L1 Mandarin and L1 Japanese users of L2 English at
intermediate levels of L2 proficiency were observed to use more iconic
gestures in their L2 than their L1, although such gestures did not seem to
compensate for missing elements in speech, particularly manner of motion
(Brown 2015).
In her review of gesture frequency among bilinguals, Nicoladis (2007,
p. 447) states that ‘there are no consistent findings across all studies (or
even most studies) relating language proficiency to gesture use’, although
‘there is a reasonably consistent finding across many studies that bilingual
adults use deictic gestures more often with their weaker language’. The
inconsistency in findings may in part be due to the different populations
observed, which include adult and child bilinguals as well as those tradi-
tionally labelled as adult second language learners. Nicoladis et al. (2007)
also point to the effects of task difficulty and gender since, in their sample,
women produced longer stories and gestured more in the L2. An additional
factor related to methodology may be data selection, and Gullberg (2011)
notes the need to examine only fluent, first descriptions in order to avoid
gestures that may be consciously deployed by the speaker for strategic
purposes later in the discourse. She also points to the importance of con-
sistency in the unit of analysis, especially in calculations of gesture rate,
although one could assume that multiple studies of gesture rate coming
from the same research teams would use comparable methods of analysis.
Underlying investigations of the frequency of gesture types in multi-
lingual versus monolingual populations is generally the question of
the extent to which gestures serve compensatory functions for multi-
competent language users. Nicoladis (2007) discusses whether, on the
one hand, bilinguals may gesture more frequently in the stronger lan-
guage since communication is likely more complex. On the other hand,
bilinguals might use more gestures in the weaker language in order to
facilitate lexical access given the nature of cross-linguistic activation and
competition (e.g. Runnqvist et al. 2012). Here an important distinction can
be made between ‘compensation’ largely for the benefit of the listener
and ‘lightening cognitive load’ for the benefit of the speaker, which
differentiates between the occurrence of gestures in place of spoken
elements and gestural facilitation in the selection of lexical units
among the multiple choices afforded to multilinguals. Despite differing
findings on the use of iconics, most research seems to support the notion
that iconic gestures are not employed for lexical compensation in an L2,
although iconics as well as other gesture types may assist L2 production
in other ways. Interestingly, Nicoladis (2007) argues that, if this is the
case, there are implications for pedagogy such that language teachers

290 AMANDA BROWN
should encourage language learners to gesture in their L2 to facilitate

production for their own benefit (see e.g. Tellier 2008).
13.3.2 Timing
While most of the research above on the temporal alignment of gesture
with speech appears to suggest the existence of cross-linguistic interac-
tions, one finding stands out as an anomaly. In their investigation of
descriptions of motion, Kellerman and van Hoof (2003) observed L1
Dutch speakers of L2 English synchronizing gestures depicting path of
motion with verbs in the L2, a pattern characteristic of neither the source
nor target languages. Since transfer from the L1 could not explain the
pattern, the authors suggested the existence of a universal interlanguage
structure in this area. This hypothesis could challenge the interpretation
of comparable findings based on gesture–speech alignment among L1
Spanish speakers of L2 English (e.g. Negueruela et al. 2004; Stam 2006);
however, the addition of corroborating data from L1 English speakers of
L2 Spanish also implying cross-linguistic influences (Negueruela et al.
2004) renders the area of temporal gesture–speech alignment in multi-
competent populations one for further investigation.
13.3.3 Semantic composition

Like Stam (2010) above, Özyürek (2002) claimed that only at an advanced
level of L2 proficiency, potentially with extensive immersion experience,
could L1 Turkish speakers of L2 English conflate information about both
manner and path of motion in single gestures (accompanying single
predicates in speech) like native speakers of English. Özyürek did not
appeal to L1 transfer as an explanation for these results, but instead
argued that the conflated constructions in speech and corresponding
gestures may be challenging to acquire, which implied a developmental
account. Interestingly, although no analyses of native speaker speech or
inferential statistical analyses were reported in the paper, native speak-
ers of Turkish, the source language, produced descriptively more con-
flated manner–path gestures than did the L2 users, a fact that escaped
notice, but would support a developmental account of the L2 results.
Further support comes from the attested L2 phenomenon of segmenta-
tion of information across separate gestures (Stam 2006).
13.3.4 Discussion of language-general patterns from the

perspective of multi-competence
There is evidence of considerable cross-cultural and cross-linguistic inter-
actions in gesture use by multi-competent populations, but there is also
evidence of features characterizing multi-competent gestures irrespective

of the languages involved. The majority of studies have considered the

frequency with which gestures are produced by multi-competent versus
mono-competent speakers and in the dominant versus non-dominant
languages. While these remain areas of controversy, there is some con-
sensus that deictic gestures appear to be used more often in second lan-
guages, particularly at lower proficiency levels, which has prompted much
debate about the extent to which gestures compensate for weaknesses in
L2 performance. The question has been reformulated to some extent,
distinguishing between compensation for the listener and facilitation for
the speaker, with the latter process seeming to account for more of the
data. For example, in their demonstration of a negative correlation
between working memory and iconic gestures among French–English
bilingual adults, Smithson and Nicoladis (2013) argued that gesture pro-
duction might facilitate storage and manipulation of language or facilitate
‘the appropriate allocation of attentional resources to ensure that the
relevant language was being spoken’ (p. 9). A reformulation away from
compensation and towards facilitation is more in line with the multi-
competence framework, which seeks to understand multilingual systems
in their own right and not from the perspective of deviant or deficient
mono-competent systems.
Very few of the studies cited in this section have simultaneously con-
sidered the L1, and there is little agreement on what the multi-competent
L1 profile looks like. This is perhaps unsurprising, since gestures are
known to vary so widely across cultures and languages that patterns
common to all L1s are likely to be rare. Moreover, in addition to cross-
cultural and cross-linguistic variation, there is also considerable individual
variability, which is rarely mentioned. Indeed, the correlations found
between gesture use and L2 proficiency in Nagpal, Nicoladis and
Marentette (2011) were extended to correlations between the L1 and L2
within individuals, indicating that each individual’s communicative style
may have a crucial role to play. Comparable gesture rates in an individual’s
L1 and L2 can be considered cross-linguistic transfer, but of a very different
sort to that normally discussed. This study underscores the importance of
including the L1 in all analyses of L2 gesture as well as the need for within-
participant correlation analyses, and highlights the individual variability
in multi-competence (see de Bot, Chapter 6, this volume).
In principle, the multi-competence framework advises against using
monolingual competence as its framework of reference. Although all
investigations of L2 gesture have recruited native speakers, often mono-
linguals, to serve as the comparison group, the research cited above, which
does not appeal to negative transfer from the L1, may be interpreted as in
line with the multi-competence framework, at least in part, in that the
studies reveal learner behaviour as distinctive and qualitatively different
from both the source and target languages. However, these studies depart
from the framework when they assume, as they often do, even implicitly,

292 AMANDA BROWN
that L2 behaviour is developmental and should, both in a predictive and

prescriptive sense, move towards the native speaker target over time.
13.4 Conclusion
Characterizations of the multi-competent system must include consid-

eration of gesture. Of relevance is how the multi-competent speaker
develops the gestural repertoires of each of the languages known and
how gestures can be employed to assist acquisition. However, gesture
analyses are also valuable in terms of the insights they may offer on
multi-competent processing and production of language, which consti-
tute the focus of this chapter.
The few qualitative studies of L2 gesture reviewed here, i.e. those with-
out the (quasi) experimental comparisons against baseline groups, gener-
ally portray L2 performance neutrally, without pejorative commentary,
and are arguably fully in line with the multi-competence framework. The
same cannot entirely be said of the majority of L2 gesture research,
although in most instances, minor adjustments in terminology could
align existing research with the framework.
One line of gesture work, which reveals characteristics of the L2 that do
not resemble those in the source and target languages, could be inter-
preted from the perspective of multi-competence as long as researchers
do not assume that such characteristics are somehow exceptional (see Vaid
and Meuter, Chapter 4, this volume) and simply part of an ‘interlanguage’,
which in turn promotes the idea of an appropriate and even inevitable
shift towards a monolingual, native speaker target. Here, the term ‘L2 user
variety’ would perhaps be a more impartial label without such assump-
tions. An additional line of research concerns itself with how features of
the L2 are influenced by features of the L1. Gullberg (2011) states that,
although they focus on evidence of L1 transfer, studies of L2 gesture,
particularly in the domain of motion, also generally reveal some level of
restructuring towards an L2 target. However, depending on their theore-
tical perspective, researchers often describe L2 performance in terms of
what L2 users ‘cannot do’, yielding characterizations of L2 performance as
‘non-targetlike’, which reflects an inherently native speaker bias. Instead,
they might focus on what L2 users ‘can do’, which would be more consis-
tent with multi-competence.
In addition to avoidance of the native speaker bias, the multi-competence
framework argues that the L1 of the multi-competent speaker should be
examined as well as their L2 and any other languages. Yet even though
comparisons of both L1 and L2 performance from multilingual speakers
versus (functional) monolinguals may help to characterize the multi-
competent system, they pose challenges when it comes to evaluation.
Documentation of the unique nature of the L1 of a multi-competent

speaker calls into question the validity of the native speaker target for L2
performance, particularly if that target comes from a monolingual
speaker of the target language. As discussed in Chapter 1 of this volume,
multi-competence does not describe a static system, but acknowledges
the development of language in any direction (e.g. expansion or contrac-
tion). However, the framework also posits that performance at any stage
should be viewed in its own right, which implies that comparison with
any other group is neither necessary nor desirable. Yet evaluation is a
necessary reality in many, though not all contexts of multilingualism, as
much desired by second language learners, users and practitioners as by
second language researchers. And any evaluation needs norms. Davies
(2009, p. 86) states, ‘A language test requires a description of the language
under test to establish the criterion against which to measure perfor-
mance.’ As argued by Brown (2013), when it comes to benchmarking, the
most valid comparison group for multi-competent speakers are other,
albeit more advanced, multi-competent speakers, ideally with compar-
able linguistic profiles with respect to the languages known. Multi-
competent benchmarking, which acknowledges that L2 performance
cannot and should not be expected to match a monolingual, native
speaker target, no longer subscribes to a deficit view of second language
acquisition. Empirical work on assessment of L2 gesture from a multi-
competence perspective, however, is non-existent and remains an area
for investigation (although see Jungheim 2001 for discussion of gesture
and assessment).
In conclusion, gesture analyses, specifically in the areas of bi-directional,
cross-linguistic influences and a multi-competent user variety, have pro-
vided some important evidence in support of the multi-competence frame-
work, but there is much left to be done. In addition to the need for much
replication work to tease apart language-specific from universal features
and a standardized inclusion of the multi-competent L1 as more than just a
baseline but an object of investigation in and of itself, longitudinal research
is desperately needed to order to reveal more about ‘L2 users themselves’
without the ‘vast quantities of information about their differences and
similarities compared to native speakers’ (Cook, Chapter 1, this volume,
p. 15). To this end, the inclusion of gesture analyses may widen the lens
through which multi-competence in action may be observed.
References
Birdsong, D. 2005. ‘Nativelikeness and non-nativelikeness in L2A

Brown, A. 2008. ‘Gesture viewpoint in Japanese and English: cross-
linguistic interactions between two languages in one speaker’,
Gesture 8 (2), 256–276.

294 AMANDA BROWN

Language Assessment Quarterly 10 (2), 219–235.
Brown, A. 2015. ‘Universal development and L1–L2 convergence in bilin-
gual construal of manner in speech and gesture in Mandarin, Japanese,
and English’, Modern Language Journal 99 (1), 66–82.
Brown, A. and Gullberg, M. 2008. ‘Bidirectional crosslinguistic influ-
ence in L1–L2 encoding of manner in speech and gesture: a study of
Japanese speakers of English’, Studies in Second Language Acquisition
30 (2), 225–251.
Choi, S. and Lantolf, J. P. 2008. ‘Representation and embodiment of mean-
ing in L2 communication’, Studies in Second Language Acquisition 30 (2),
191–224.
Davies, A. 2009. ‘Assessing World Englishes’, Annual Review of Applied
Efron, D. 1941/1972. Gestures, Race and Culture. The Hague: Mouton.
Freleng, F. 1950. Canary Row (film, animated cartoon). New York: Time
Warner.
Gabrys-Barker, D. (ed.) 2012. Cross-linguistic Influences in Multilingual Language
Acquisition. New York: Springer.
Goldin-Meadow, S. 1998. ‘The development of gesture and speech as an
integrated system’, in J. M. Iverson and S. Goldin-Meadow (eds), The
Nature and Functions of Gesture in Children’s Communication, pp. 29–42. San
Francisco: Jossey-Bass.
Gullberg, M. 1998. Gesture as a Communication Strategy in Second Language
Discourse: A Study of Learners of French and Swedish. Lund University Press.
Gullberg, M. 2003. ‘Gestures, referents, and anaphoric linkage in learner
varieties’, in C. Dimroth and M. Starren (eds), Information Structure,
Linguistic Structure and the Dynamics of Language Acquisition, pp. 311–328.
Amsterdam: Benjamins.
Gullberg, M. 2006a. ‘Some reasons for studying gesture and second lan-
guage acquisition (Hommage à Adam Kendon)’, International Review of
Applied Linguistics 44 (4), 103–124.
Gullberg, M. 2006b. ‘Handling discourse: gestures, reference tracking, and
communication strategies in early L2’, Language Learning 56 (1), 155–196.
Gullberg, M. 2008. ‘Gestures and second language acquisition’, in
P. Robinson and N. C. Ellis (eds), Handbook of Cognitive Linguistics and
Second Language Acquisition, pp. 276–305. New York: Routledge.
Gullberg, M. 2009a. ‘Gestures and the development of semantic represen-
tations in first and second language acquisition’, Languages, Interaction
and Acquisition 1, 117–139.
Gullberg, M. 2009b. ‘Reconstructing verb meaning in a second language:
how English speakers of L2 Dutch talk and gesture about placement’,
Annual Review of Cognitive Linguistics 7, 222–245.

Gullberg, M. 2011. ‘Thinking, speaking and gesturing about motion in

more than one language’, in A. Pavlenko (ed.), Thinking and Speaking in
Two Languages, pp. 143–169. Bristol: Multilingual Matters.
Gullberg, M. 2012a. ‘Bilingualism and gesture’, in T. K. Bhatia and
W. C. Ritchie (eds), The Handbook of Bilingualism and Multilingualism,
pp. 417–437. Wiley-Blackwell.
Gullberg, M. 2012b. ‘Gesture analysis in second language acquisition’, in
M. Gullberg and J. Williams (eds), Cognitive Second Language Acquisition (The
Encyclopedia of Applied Linguistics). Blackwell.
Jarvis, S. 2010. ‘Comparison-based and detection-based approaches to trans-
fer research’, in L. Roberts, M. Howard, M. Ó Laoire and D. Singleton (eds),
EUROSLA Yearbook 10, pp. 169–192. Amsterdam: John Benjamins.
Johnson, H. G., Ekman, P. and Friesen, W. V. 1981. ‘Communicative body
movements: American emblems’, in A. Kendon (ed.), Nonverbal
Communication, Interaction, and Gesture, pp. 401–419. The Hague: Mouton.
Jungheim, N. O. 2001. ‘The unspoken element of communicative compe-
tence: evaluating language learners’ non-verbal behavior’, in T. Hudson
and J. D. Brown (eds), A Focus on Language Test Development: Expanding the
Language Proficiency Construct across a Variety of Tests, pp. 1–34. Honolulu:
University of Hawai’i, Second Language Teaching and Curriculum Center.
Second Language Acquisition. Oxford: Pergamon.
Kellerman, E. and van Hoof, A.-M. 2003. ‘Manual accents’, International
Review of Applied Linguistics 41 (3), 251–269.
Kendon, A. 1972. ‘Some relationships between body motion and speech: an
analysis of an example’, in A. W. Siegman and B. Pope (eds), Studies in
Dyadic Communication, pp. 177–210. New York: Pergamon.
Kendon, A. 1980. ‘Gesticulation and speech: two aspects of the process of
utterance’, in M. R. Key (ed.), The Relationship of Verbal and Nonverbal
Communication, pp. 207–227. The Hague: Mouton.
Kendon, A. 1986. ‘Some reasons for studying gesture’, Semiotica 62 (1/2),
3–28.
Kendon, A. 2004. Gesture: Visible Action as Utterance. Cambridge University
Press.
Kita, S. 2000. ‘How representational gestures help speaking’, in D. McNeill
(ed.), Gesture and Language: Window into Thought and Action, pp. 162–185.
Cambridge: Cambridge University Press.
Kita, S. (ed.) 2003. Pointing: Where Language, Culture and Cognition Meet.
Kita, S. and Özyürek, A. 2003. ‘What does cross-linguistic variation in
semantic coordination of speech and gesture reveal? Evidence for an
interface representation of spatial thinking and speaking’, Journal of
Memory and Language 48 (1), 16–32.
McNeill, D. 1992. Hand and Mind: What the Hands Reveal about Thought.
Chicago University Press.

296 AMANDA BROWN
McNeill, D. 2001. ‘Imagery in motion event descriptions: gestures as part of

thinking-for-speaking in three languages’, in Proceedings of the Twenty-
Third Annual Meeting of the Berkeley Linguistics Society, pp. 255–267.
McNeill, D., Levy E. T. and Pedelty, L. L. 1990. ‘Speech and gesture’, in
G. R. Hammond (ed.), Cerebral Control of Speech and Limb Movements,
pp. 203–256. Amsterdam: North-Holland.
Morris, D. 1977. Manwatching. New York: Abrahams.
Morris, D., Collett, P., Marsh, P. and O’Shaughnessy, M. 1979. Gestures: Their
Origin and Distribution. London: Jonathan Cape.
Müller, Cornelia 2001. ‘Gesture-space and culture’, in C. Cavé, I. Guaı̈tella
and S. Santi (eds), Oralité et Gestualité: Interactions et Comportements
Multimodaux dans la Communication (Actes du Colloque ORAGE 2001, Aix-
en-Provence, 18–22 June 2001), pp. 565–571. Paris: l’Harmattan.
Nagpal, J., Nicoladis, E. and Marentette, P. F. 2011. ‘Does proficiency or task
difficulty explain bilinguals’ gesture?’ International Journal of Bilingualism
15, 205–214.
Negueruela, E., Lantolf, J. P., Jordan, S. R. and Gelabert, J. 2004. ‘The
“private function” of gesture in second language speaking activity: a
study of motion verbs and gesturing in English and Spanish’,
International Journal of Applied Linguistics 14 (1), 113–147.
Nicoladis, E. 2007. ‘The effect of bilingualism on the use of manual ges-
tures’, Applied Psycholinguistics 28 (3), 441–454.
Nicoladis, E., Pika, S. and Marentette, P. F. 2009. ‘Do French–English bilin-
gual children gesture more than monolingual children?’, Journal of
Psycholinguistic Research 38, 573–585.
Nicoladis, E., Pika, S., Yin, H. and Marentette, P. F. 2007. ‘Gesture use in
story recall by Chinese–English bilinguals’, Applied Psycholinguistics 28,
721–735.
Odlin, T. 2008. ‘Conceptual transfer and meaning extensions’, in P. Robinson
and N. Ellis (eds), Handbook of Cognitive Linguistics and Second Language
Acquisition, pp. 306–340. New York: Routledge.
Özyürek, A. 2002. ‘Speech–gesture relationships across languages and in
second language learners: implications for spatial thinking for speak-
ing’, in B. Skarabela, S. Fish and A. H.-J. Do (eds), Proceedings of the 26th
Annual Boston University Conference on Language Development, pp. 500–509.
Boston, MA: Cascadilla Press.
Pika, S., Nicoladis, E. and Marentette, P. F. 2006. ‘A cross-cultural study on
the use of gestures: evidence for cross-linguistic transfer?’, Bilingualism:
Runnqvist, E., FitzPatrick, I., Strijkers, K. and Costa, A. 2012. ‘An appraisal of
the bilingual language production system: quantitatively or qualitatively
different from monolinguals?’, in T. Bhatia and W. C. Ritchie (eds), The
Handbook of Bilingualism and Multilingualism, pp. 244–266. Wiley-Blackwell.
Santos, P. 1974. ‘Examples of figurative gestures which characterize the
Brazilian people’, American Foreign Language Teacher 4 (4), 22–25.

Schegloff, E. A. 1984. ‘On some gestures’ relation to talk’, in J. M. Atkinson

and J. Heritage (eds), Structures of Social Action, pp. 266–296. Cambridge
University Press.
Sherman, J. and Nicoladis, E. 2004. ‘Gestures by advanced Spanish–English
second-language learners’, Gesture 4 (2), 143–156.
Slobin, D. I. 1996. ‘From “thought and language” to “thinking for speak-
ing”’, in J. J. Gumperz and S. C. Levinson (eds), Rethinking Linguistic Relativity,
Smithson, L. and Nicoladis, E. 2013. ‘Verbal memory resources predict
iconic gesture use among monolinguals and bilinguals’, Bilingualism:
Smithson, L., Nicoladis, E. and Marentette, P. 2011. ‘Bilingual children’s
gesture use’, Gesture 11 (3), 330–347.
So, W.-C. 2010. ‘Cross-cultural transfer in gesture frequency in Chinese–
English bilinguals’, Language and Cognitive Processes 25 (10), 1335–1353.
Sparhawk, C. M. 1981. ‘Contrastive-identificational features of Persian
gesture’, in A. Kendon (ed.), Nonverbal Communication, Interaction and
Gesture, pp. 421–458. The Hague: Mouton.
Stam, G. 2006. ‘Thinking for speaking about motion: L1 and L2 speech and
gesture’, International Review of Applied Linguistics 44 (2), 143–169.
Stam, G. 2010. ‘Can an L2 speaker’s patterns of thinking for speaking
change?’, in Z. Han and T. Cadierno (eds), Linguistic Relativity in SLA:
Thinking for Speaking, pp. 59–83. Clevedon: Multilingual Matters.
Talmy, L. 2000. Toward a Cognitive Semantics, vols I and II. Cambridge, MA:
MIT Press.
Tellier, M. 2008. ‘The effect of gestures on second language memorisation
by young children’, Gesture 8 (2), 218–235.
Von Stutterheim, C., Nüse, R. and Murcia-Serra, J. 2002. ‘Cross-linguistic
differences in the conceptualisation of events’, in H. Hasselgaº rd,
S. Johansson, B. Behrens and C. Fabricius–Hansen (eds), Information
Structure in a Cross-linguistic Perspective, pp. 179–198. Amsterdam: Rodopi.
Von Stutterheim, C. and Nüse, R. 2003. ‘Processes of conceptualization in
language production: language-specific perspectives and event con-
strual’, Linguistics 41 (5), 851–881.
Yoshioka, K. and Kellerman, E. 2006. ‘Gestural introduction of Ground
reference in L2 narrative discourse’, International Review of Applied
Linguistics 44 (2), 171–193.

14
Pragmatic transfer in
foreign language learners:
a multi-competence
perspective
I-Ru Su
In recent years there has been a growing number of second language

acquisition researchers probing the issue of language transfer (or cross-
linguistic influence) from a multi-competence perspective (e.g. Brown and
Gullberg 2011; Pavlenko and Jarvis 2002; Su 2010; Zareva 2010). Multi-
competence, a concept proposed by Cook (1991), refers to the knowledge
of two or more languages in one mind. In contrast to the traditional view of
language transfer, which emphasizes uni-directional influence from L1 to
L2 (e.g. Gass and Selinker 1992), the multi-competence view assumes that
transfer can go in two directions and there is a complex interaction
between the L2 user’s language systems.
Evidence supporting the multi-competence view of second language acqui-
sition mainly comes from two strands of research: research examining the
effects of the L2 on the L1 (e.g. Bylund and Jarvis 2011; papers in Cook 2003;
Dussias and Sagarra 2007) and the relationships between the L2 user’s lan-
guage systems (e.g. Brown and Gullberg 2011; Mennen 2004; Pavlenko and
Jarvis 2002; Su 2001). Studies of L2 influence on the L1 have shown that L2
users’ knowledge and use of their L1 is distinguishable from that of L1
monolinguals. For example, Laufer (2003) found that the collocational knowl-
edge of Russian immigrants in Israel differed from that of Russian monolin-
guals. The Russian immigrants tended to use L2 knowledge to judge L1 word
collocations. For instance, they recognized close the telephone as a correct
collocation, a typical usage in Hebrew but regarded as incorrect in Russian.
In a study of sentence parsing, Dussias and Sagarra (2007) also demonstrated
that Spanish–English bilinguals with extensive L2 immersion experience
applied L2-specific strategies in processing L1 ambiguous sentences.
Investigations into the relationships of the L2 user’s language systems
have evidenced complex patterns of interaction between the languages in

Pragmatic transfer in foreign language learners 299
the L2 user’s mind. Su (2001), for instance, explored the L1–L2 transfer of
sentence processing strategies among L2 users of English and Chinese. Her
results showed that L2 users at all proficiency levels carried their L1
strategies over to L2 processing, but at the same time the L2 users of higher
proficiency applied the L2 strategies in L1 processing. Moreover, within-
group comparisons indicated that the advanced L2 users seemed to have
merged the processing strategies of their two languages and applied the
combined strategies in processing both languages. Su’s study indicated
that the interaction patterns of the L2 user’s two language systems might
change as a function of proficiency. In a more recent study, Brown and
Gullberg (2011) investigated bi-directional transfer in the expressions of
motion events among Japanese learners of English, focusing on the differ-
ent path components. Their results also indicated a bi-directional interac-
tion between the learners’ two linguistic systems. They found a strong L1
influence on the learners’ expressions of path in the L2 as well as a subtle
trace of L2 influence on the learners’ L1 productions. Bi-directional trans-
fer is also observed in other linguistic areas such as segmental phonology
(Zampini and Green 2001), intonation patterns (Mennen 2004), morpho-
syntactic structures (Pavlenko and Jarvis 2002), word association (Zareva
2010), and narrative structure and rhetoric (Stavans 2003). These studies
converged on the findings that cross-linguistic influence works both ways,
from L1 to L2 and from L2 to L1, and that the L2 user’s linguistic represen-
tations of both languages are qualitatively distinct from those of mono-
linguals of either language. These findings attest to Cook’s claim that L2
users are unique in their own right. The framework of multi-competence
thus opens up new perspectives in second language acquisition research;
that is, comparisons between L2 users and native speakers should shift
from sticking to L2 deficiencies to discovering their unique characteristics
(Cook 2003).
Most of the research that examined the linguistic representations of the
multi-competence of the L2 user has focused on the L2 user’s grammatical
competence. Relatively less research has addressed the L2 user’s commu-
nicative competence, that is, the ability to speak appropriately in a given
language community. Sociocultural rules for appropriate speaking vary
from one speech community to another (Hymes 1964). In order to com-
municate successfully with native speakers of the target language, L2 users
need to be equipped not only with linguistic knowledge but also with
knowledge of sociocultural rules for appropriate use. To date only a
small number of studies have explored L2 users’ pragmatic competences
in their two languages, and they mainly concern immigrants who had
lived in the target language country for an extended period of time
(Blum-Kulka 1990; Valdés and Pino 1981). Valdés and Pino (1981) looked
at how Mexican-American bilinguals responded to compliments in their L1
and L2 and compared their compliment responses to those used by mono-
linguals. They found the bilinguals had merged the compliment response

300 I-RU SU
sets of their two languages and behaved differently from monolinguals.

Blum-Kulka (1990) examined the requesting behaviour of American immi-
grants in Israel and found that the request realizations of these American
immigrants presented a level of directness that can be situated in between
American and Israeli requests.
Speech act behaviours are culture-sensitive, and so immersion experi-
ence may be required to fully acquire the L2 sociocultural rules to speak
appropriately (Yu 2004) and thus possibly to trigger restructuring in the L1.
One may wonder if foreign language learners, who learn and use the L2 in
a non-native environment, would evidence change in their L1 pragmatic
competence. In this chapter, I will review two studies that probed
bi-directional transfer in the pragmatic competences of foreign language
learners within the framework of multi-competence. Su (2010, 2012)
recruited Chinese-speaking learners of English from Taiwan and examined
their requesting and apologizing behaviour, respectively, in their two
languages. The studies aimed to explore how the two pragmatic systems
interact in the foreign learner’s mind. In addition, in comparison with
previous studies that examined L1–L2 interactions in L2 users immersed in
the target language environment, the two studies addressed whether
language learning/use context affects the way the L2 learner’s/user’s two
language systems are integrated.
Before we move on to the discussion of Chinese EFL learners’ requesting
and apologizing behaviours, cross-cultural comparisons of the two speech
act behaviours in English and Chinese are in order.
14.1 Cross-cultural comparisons of requesting

and apologizing behaviour in Chinese and English
According to Brown and Levinson (1987), both request and apology are
face-threatening acts (FTAs) that pose a threat to the hearer’s negative face,
that is, freedom from imposition. People adopt various strategies to mini-
mize or eliminate such threats, and more polite strategies are adopted by
the speaker if the FTA is deemed serious. Brown and Levinson suggested
that the weightiness of FTAs is influenced by social or contextual factors,
such as relative power and distance between the interlocutors and the
ranking of the imposition, and that these social variables are culturally
sensitive. Indeed, cross-cultural research on requesting and apologizing
behaviours has shown that although L1 speakers from different languages
adopt a similar set of strategies, they respond differently to these social
factors (e.g. Barnlund and Yoshioka 1990; Bergman and Kasper 1993; Blum-
Kulka et al. 1989; Yu 1999).
Previous studies on Chinese and English requests have indicated that
both Chinese and English native speakers prefer conventionally indirect
strategies (‘Could you clean up the room?’) to other major request

strategies, but Chinese native speakers employ more direct strategies

(‘Clean up the room!’) than English native speakers do in certain social
contexts (Chen 2006; Lee-Wang 1994; Yu 1999). Chinese native speakers
are more direct when formulating requests in the situations where the
speaker is in a position of higher status or authority, where the speaker is
socially close to the addressee, or where the request is deemed to be easily
carried out (Lee-Wang 1994). The observation that Chinese speakers are
more direct than English speakers in making requests in the aforemen-
tioned situations can be accounted for by reference to cultural differences.
Compared to most Western cultures, many Asian cultures put more empha-
sis on the relational hierarchy within the family and society and, therefore,
direct strategies are considered more socially appropriate than indirect ones
in the aforementioned contexts. Being indirect in these situations can be
perceived as inappropriate in Chinese society (Lee-Wang 1994).
With regard to the speech act of apology, previous studies have also
shown that while Chinese and English native speakers adopt a similar set
of strategies, the former are more sensitive than the latter to the relative
social status of the interlocutor. While Chinese native speakers utilize more
elaborated strategies when apologizing to a superior than to an inferior (Tsai
2002), English native speakers do not vary their strategy use according to the
interlocutor’s relative status (Barnlund and Yoshioka 1990; Bergman and
Kasper 1993). The cross-cultural difference in terms of relational hierarchy
within the family and society can again explain why Chinese and English
native speakers react differently to the social factor of status when making
apologies. Chinese native speakers may consider the damage to the face of
the hearer who is of higher status to be a more serious FTA than Westerners
do and thus adopt more elaborated apology strategies to mitigate the impact
of an offensive act.
Given the differences between Chinese and English in the speech act
performance of request and apology, one may wonder how Chinese EFL
learners behave in their two languages. Let’s start with their requesting
behaviour.
14.2 Chinese EFL learners’ requesting behaviour

in L1 and L2
14.2.1 Participants
The study consisted of 120 college students. They comprised four groups:
two groups of Chinese EFL learners varying in their proficiency level and
two groups of native controls (i.e. Chinese and English) who served as a
baseline against which L2 learners’ speech act performance was compared
to identify possible cross-linguistic influence from both directions. There
were 30 participants in each group. The English native controls were
recruited from a university in the USA, and they neither knew Chinese

302 I-RU SU
nor were fluent in any foreign language. The Chinese-speaking partici-

pants were recruited from Taiwan. Chinese native speakers who did not
know any English at all were hard to find, as English is a mandatory subject
in Taiwan in junior and senior high school education. One solution sug-
gested by Cook (2003) is to find L2 learners with very limited proficiency in
English and contrast them with those of higher English proficiency.
Hence, a TOEFL test (paper version) was administered to the Chinese-
speaking subjects and they were placed into one of the following groups
according to their scores: Chinese native controls (average 350); intermedi-
ate (average 470), and advanced (above 600). The cut-off points were arbi-
trarily set with the intention of enlarging the gap in English proficiency
among different groups of Chinese-speaking subjects so that possible
differences among groups in their requesting and apologizing behaviour
could be observed.
14.2.2 Instrument and procedure

Written discourse completion tests (DCT) were used as an instrument to
collect the participants’ responses. The questionnaire contained 15 situa-
tions, including 9 request situations and 6 apology scenarios. Two different
social situations were included in order to prevent participants from
answering mechanically and to collect more data on L2 learners’ speech
act performance. The results of the EFL learners’ apologizing performance
will be reported in the next section.
Given that the relative social status and distance between the speaker
and the hearer have been shown to affect people’s speech act behaviours,
the request and apology situations were designed to vary along these two
social parameters. In the request scenarios, the variable of social status was
categorized into three levels: high (the requester has a higher status than
the requestee), equal (the requester has an equal status to the requestee),
and low (the requester has a lower status than the requestee). The variable
of social distance varied on three levels: close, acquaintance, and stranger.
The request situations in the DCT included the following:
S1 Loud music. The person asks his/her younger sister (or brother) to turn
down music.
S2 Work overtime. The manager asks his secretary to work on Saturday.
S3 Illegal parking. The police officer asks the driver to move his/her car.
S4 Computer crash. The person asks his/her roommate to fix his/her computer.
S5 Free ride. The person asks his/her neighbour to give him/her a ride
home.
S6 Book searching. The person asks a tall person nearby to help get a book
on a higher bookshelf.
S7 Money loan. The person asks his/her father to lend him/her money to
buy a computer.

S8 Late homework. The student asks the professor to extend the deadline of
a term paper.
S9 Job searching. The person calls to ask about a part-time job.
Below is an example of a request situation:
You are studying in your room for a coming exam, but you can’t concen-
trate because your younger brother (or sister) is listening to loud music in
the next room. If you asked him (or her) to use his (or her) headphones,
what would you say?
You say, ‘____________________________’
Both English and Chinese versions of the questionnaire were created. The
native controls completed one DCT in their respective native language,
while Chinese EFL learners filled out the English and Chinese DCTs on
different days with an interval of two weeks or longer; this was to mini-
mize any memory and direct translation effects. The administration of the
two tests in the EFL groups was counterbalanced to avoid the task effect;
half of the EFL learners in each group did the English DCT first and the
Chinese DCT second, and vice versa for the other half.
14.2.3 Data analysis

Participants’ request realizations were coded according to the coding
scheme for requests developed by Cross-Cultural Speech Act Realization
Project (Blum-Kulka et al. 1989). Following CCSARP, the head act of the
request (i.e. the request utterance proper) is distinguished from the strate-
gies by which it can be modified. Such strategies can be applied internally
or externally to the head act. Thus, the participants’ request realizations
were analysed in terms of three aspects: request strategies (i.e. the direct-
ness level of the request proper), internal modification (i.e. lexical/phrasal
downgraders), and external modification (i.e. supportive moves). There are
three major request strategies which vary in the directness level, namely
directives (‘Turn down the music!’), conventional indirectives (‘Could you
turn down the music?’), and non-conventional indirectives (‘The music is
too loud!’). Previous studies have reported that among the three major
strategies, conventionally indirect strategies are used most often by English
and Chinese native speakers in making requests, and so the analysis of
request strategies in the study centred on conventional indirectives.
Lexical/phrasal downgraders and supportive moves are both optional
additions to the head act to mitigate the impositive force of requests.
Lexical/phrasal downgraders soften the requestive force by modifying the
head act internally through specific lexical and phrasal choices, such as
politeness markers (‘please’), consultative devices (‘I was wondering if . . .’),
subjectivizer (‘I’m afraid’), and so on. Supportive moves modify the head act
externally. They include justifications, disarmers, promises of reward, and

304 I-RU SU
Table 14.1 Examples of supportive moves identified in the study (from Su

2010, p. 92)
Preparatory Work overtime. Are you free on Saturday?

Pre-commitment Work overtime. I’d like to ask you for a favour.
Grounder Work overtime. We have been very busy lately.
Apology Loud music. I’m sorry.
Thanking Work overtime. I’d appreciate it.
Promise Late homework. I promise I’ll do my best and finish it as soon as
possible.
Promise of reward Work overtime. I’ll be willing to pay you overtime if you are up to it.
Cost minimizer Money loan. I’ll pay you back right away.
Direct appeal Free ride. I really need your help in this bad weather!
Sweetener Work overtime. Ms Li, you are the most capable staff in our company.
Disarmer Work overtime. Ms Li, I know Saturday is normally your day off.
so forth. Table 14.1 presents the types of supportive moves identified in the
study. Two fluent Chinese–English bilinguals coded the data independently.
The inter-rater agreement was 95%. Consensus was reached when there was
disagreement.
The two native control groups were compared first, to provide the base-
line data for the examination of L1 and L2 transfer. In order to determine
whether transfer was operative, the study adopted an operational defini-
tion of language transfer based on Selinker (1969). According to this
definition, forward transfer (i.e. transfer from L1 to L2) obtained when
there were statistically significant differences in the frequencies of use of a
pragmatic feature between L1 and L2 native controls’ request realizations,
and between learners’ L2 request performance and that of L2 native con-
trols. On the other hand, backward transfer (i.e. transfer from L2 to L1) held
when statistically significant differences in the frequencies of use of a
pragmatic feature were found between L1 and L2 native controls’ request-
ing behaviour, and between learners’ L1 request performance and that of
L1 native controls. The same criteria for determining transfer were also
applied in the study of apology to be reviewed in the next section.
14.2.4 Results and discussion

English and Chinese native controls’ requesting behaviour
Results of independent t-tests indicated that English native speakers
(ENSs) adopted significantly more conventionally indirect strategies
than Chinese native speakers (CNSs) (t = 8.99, p = 0.000); however, they
did not vary significantly from each other in the frequency usage of
lexical/phrasal downgraders or supportive moves.
Figure 14.1 displays the distribution of the three major request strate-
gies in the two native control groups. As we can see, conventionally
indirect strategies were the most favoured strategies for both ENSs and

Table 14.2 Mean number of conventionally indirect strategies in relation to

social variables in the L1 control groups (from Su 2010, p. 93). N = 30 in each
ENSs CNSs t-test

Mean SD Mean SD t p
Status
High 1.97 0.76 0.33 0.40 9.17 0.000
Equal 2.73 0.45 2.23 0.57 3.78 0.000
Low 2.53 0.57 2.13 0.68 2.46 0.017
Distance
Close 2.57 0.50 1.17 0.83 7.87 0.000
Acquaintance 2.77 0.43 1.93 0.69 5.61 0.000
Stranger 1.90 0.66 1.60 0.56 1.89 n.s.
Note. The maximum number for each social category is 3.
8
7
Mean frequency of use
6
5
4 ENSs
3 CNSs
2
1
0
Directives Conventional Non-conventional
indirectives indirectives
Figure 14.1 English and Chinese native speakers’ use of different types of request strategies
CNSs; nonetheless, CNSs appeared to use much more direct strategies than
their English counterparts in making requests.
Table 14.2 presents the control groups’ frequencies of use of conven-
tionally indirect strategies in relation to the social variables of status and
distance. As shown, CNSs utilized conventional indirectives significantly
less often than did ENSs in most of the social conditions. CNSs were most
direct when making requests to people who are of lower status or socially
close. This finding is line with the previous research on Chinese requests
(Lee-Wang 1994).
Chinese EFL learners’ requesting behaviour in their L2
Comparing Chinese EFL learners’ requesting behaviour in English to that
of the ENSs, we found that EFL learners differed from ENSs in all dimen-
sions in their request realizations (Table 14.3). Regarding request strate-
gies, the statistical analysis indicated that regardless of proficiency level,
the EFL learners utilized conventional indirectives significantly less often
than ENSs did.

306 I-RU SU
Table 14.3 English requests by ENSs and Chinese EFL learners (from Su
2010, p. 94). N = 30 per group
ENSs Inter. EFL Adv. EFL ANOVA

Mean SD Mean SD Mean SD F p
Con. indirect strategies 7.23a 0.89 6.43b 1.52 6.33b 1.21 4.76 0.01
Lex./phr. downgraders 3.73a 2.21 3.67a 1.47 2.67b 1.89 3.01 0.05
Supportive moves 5.97a 1.55 5.23a 2.08 6.53b 1.66 4.08 0.02
Note. The maximum number for each category is 9. Means with different alphabet letters
(a and b) within the same category differ significantly (p < 0.05) by the post hoc Tukey test.
8
7
6
5
ENSs
4
Inter. EFL
3
Adv. EFL
2
1
0
Figure 14.2 The frequencies of different types of request strategies among ENSs and EFL
learners
Figure 14.2 presents the frequency distributions of the three major types
of request strategies used by ENSs and the EFL learners. We can see that
overall both learner groups employed fewer conventionally indirect stra-
tegies and more direct strategies than ENSs did; the EFL learners appeared
to be more direct than ENSs in making requests.
A closer examination of the data showed that both intermediate and
advanced learners adopted significantly fewer conventional indirectives
than ENSs in the situations in which the requester is of higher status or
familiar to the requestee (Table 14.4). This indicates forward transfer (i.e.
transfer from L1 to L2) as CNSs also exhibited differences in their request-
ing behaviour from ENSs in the aforementioned situations (see Table 14.2).
With respect to internal and external modifications, the results showed
that the intermediate learners did not diverge from ENSs in these two
aspects, whereas the advanced learners employed significantly fewer
lexical/phrasal downgraders but more supportive moves than did the
ENSs and the intermediate learners (Table 14.4). Although significant
differences were observed in the use of internal and external modifiers
between the advanced EFL learners and the ENSs, they cannot be attribu-
ted to the influence of the L1 because the request realizations of English


social variables in the English DCT (from Su 2010, p. 94)

Status
High 1.97a 0.76 1.33b 0.99 1.10b 0.96 7.26 0.001
Equal 2.73 0.45 2.63 0.61 2.80 0.48 0.78 n.s.
Low 2.53 0.57 2.47 0.86 2.43 0.68 0.15 n.s.
Distance
Close 2.57a 0.5 1.93b 0.91 1.93b 0.69 7.74 0.001
Acquaintance 2.77 0.43 2.47 0.63 2.53 0.68 2.14 n.s.
Stranger 1.90 0.66 2.03 0.56 1.87 0.43 0.75 n.s.
Note. The maximum number in each social category is 3. Means with different alphabet
letters within the same category differ significantly (p < 0.05) by the post hoc Tukey test.
and Chinese native controls did not differ in those two dimensions.
Previous studies on L2 pragmatics have reported that L2 learners, particu-
larly advanced non-native speakers, tend to produce lengthy utterances
when making requests (Blum-Kulka and Olshtain 1986; Byon 2004; Cenoz
and Valencia 1996). One possible reason, as suggested by Blum-Kulka and
Olshtain (1986), is that L2 learners generally lack confidence and are eager
to get their messages across. The intermediate EFL learners, by compar-
ison, may not have sufficient linguistic facility to elaborate on the mitigat-
ing supportive moves to soften the requestive force, and so their overall
request realizations were shorter than the advanced learners’. To compen-
sate for their weaker linguistic ability to produce longer utterances, the
intermediate learners used far more lexical/phrasal downgraders such as
‘please’ to indicate politeness.
Chinese EFL learners’ requesting behaviour in their L1

The Chinese EFL learners’ request responses in Chinese were compared to
those of the Chinese native controls to see if their request performance
was distinguishable from that of CNSs as a result of learning an additional
language. As shown in Table 14.5, both intermediate and advanced lear-
ners adopted significantly more conventionally indirect strategies than
CNSs; no significant difference was found in their frequency use of inter-
nal and external modifiers.
Figure 14.3 displays the frequency distributions of the three major
request strategies among CNSs and EFL learners. As shown, both learner
groups adopted fewer directives but more conventional indirectives than
CNSs did; the EFL learners appeared to be less direct than CNSs in making
requests.
A closer look at the participants’ strategy use in different social condi-
tions revealed that both learner groups employed significantly more con-
ventional indirectives than CNSs when making requests to inferiors, and
that the advanced learners also used considerably more conventional

308 I-RU SU
Table 14.5 Chinese requests by CNSs and Chinese EFL learners (from Su
2010, p. 96). N = 30 per group
CNSs Inter. EFL Adv. EFL ANOVA

Con. indirect strategies 4.73a 1.23 5.70b 1.42 5.60b 1.28 4.94 0.009
Lex./phr. downgraders 3.43 1.69 3.77 1.99 3.33 1.60 0.49 n.s.
Supportive moves 5.67 1.67 5.80 1.65 5.93 1.36 0.22 n.s.
Note. The maximum number of each category is 9. Means with different alphabet letters
within the same category differ significantly (p < 0.05) by the post hoc Tukey test.
8
7
6
5
CNSs
4
Inter. EFL
3
Adv. EFL
2
1
0
Figure 14.3 The frequencies of different types of request strategies among CNSs and EFL
learners
indirectives when making requests to people who are socially close to

them (Table 14.6). This is an indication of backward transfer (i.e. transfer
from L2 to L1) as the English request realizations have been shown to be
more indirect than the Chinese ones, especially in the aforementioned
social situations (see Table 14.2).
Within-group comparisons in English versus Chinese

Within-group comparisons were made to see if the Chinese EFL learners
differentiated or merged the pragmatic repertoires of their two languages.
Results of paired-sample t-tests indicated that both intermediate and
advanced learners adopted significantly more conventionally indirect stra-
tegies in making requests in English than in Chinese (Tables 14.7 and 14.8).
It is also interesting to note that the advanced learners employed consider-
ably more supportive moves when making requests in English than in
Chinese. It appears that both groups of learners differentiated their two
pragmatic systems to some extent, more so for the advanced learners. To
illustrate, the responses of the same advanced learner given in ‘Work
overtime’ are provided below. In ‘Work overtime’, the power relation is
high (the manager) to low (the secretary). In the examples, we can see the


social variables in the Chinese DCT (from Su 2010, p. 96)

Status
High 0.33a 0.61 0.80b 0.76 0.83b 0.79 4.47 0.01
Equal 2.23 0.57 2.40 0.67 2.43 0.57 0.94 n.s.
Low 2.13 0.68 2.50 0.57 2.33 0.66 0.09 n.s.
Distance
Close 1.17a 0.83 1.47 1.10 1.73b 0.83 2.78 0.07#
Acquaintance 1.93a 0.69 2.27 0.64 2.37b 0.67 3.47 0.035
Stranger 1.60a 0.53 1.97b 0.41 1.50a 0.51 7.27 0.001
#
Marginally significant.
Note. The maximum number in each social category is 3. Means with different alphabet
letters within the same category differ significantly (p < 0.05) by the post hoc Tukey test.
Table 14.7 Intermediate EFL learners’ requests in English and Chinese (from
Su 2010, p. 97)
English test Chinese test t-test

Mean SD Mean SD t p
Con. indirect strategies 6.43 1.52 5.70 1.42 3.00 0.005

Lex./phr. downgraders 3.67 1.47 3.77 1.99 −0.26 n.s.
Supportive moves 5.23 2.08 5.80 1.65 −1.47 n.s.
Table 14.8 Advanced EFL learners’ requests in English and Chinese (from Su
2010, p. 98)

Mean SD Mean SD t p
Con. indirect strategies 6.33 1.21 5.60 1.28 3.34 0.002

Lex./phr. downgraders 2.67 1.89 3.33 1.60 −1.76 n.s.
Supportive moves 6.53 1.66 5.93 1.36 1.87 0.07#
#
Marginally significant.
same advanced learner used a conventionally indirect strategy in addres-

sing the secretary in English but she used a direct strategy when making
the same request in Chinese. The request proper was italicized in the
examples. The learner also used more supportive moves in her English
request than in the Chinese one.
(1) a. Ms Li, you are an excellent worker and deserve a nice break, so I
feel really sorry to ask you to continue on Saturday. But it’s
getting a bit busy here so I really need your help. Can you do me
this favour?

310 I-RU SU
b. 李小姐，這個禮拜業務比較多，所以還要麻煩你星期六來公司，辛
苦你了. li xiaojie, zhege libai yewu bijiao duo, suoyi haiyao mafan
ni xingqiliu lai gongsi, xinku ni le.
(Ms Li, this week (we) have more work to do, so (I) need to bother you
to come to the office on Saturday. Thanks.)
14.3 Chinese EFL learners’ apologizing behaviour

in L1 and L2
14.3.1 Apology situations in the DCT

Now we turn to the Chinese EFL learners’ apology performance. The
apology situations in the DCT also varied along the social factors of status
and distance. The variable of social status contained three levels: high (the
apologizer has a higher status than the apologizee), equal (the apologizer
has an equal status to the apologizee), and low (the apologizer has a lower
status than the apologizee). Different from the request situations, the
variable of social distance varied only on two levels in the apology situa-
tions: acquaintance (the interactants knew each other) and stranger (the
interactants had never met before). The offence situations in the DCT
included the following:
S1 Exam papers. The teacher promised to give back students’ exam papers
in class, but then informs the students that he/she has not finished
grading them.
S2 Job interview. The manager scheduled a job interview with an applicant,
but the manager is 20 minutes late because s/he received a phone call
before the interview.
S3 Watch movie. The person planned to watch a movie with a friend, but
s/he is late; the movie has started before s/he arrives.
S4 Car incident. The person accidentally backs into someone’s car.
S5 Return book. The student promised to return the book s/he had bor-
rowed from a professor, but s/he forgets to bring it.
S6 Coffee spill. The waiter/waitress spills some coffee on a customer’s
jacket.
Below is an example of an offense situation:
You are a teacher. After the mid-term exam, the students are anxious to
know their scores. You promised to give them back the exam papers today,
but you haven’t finished grading them. What would you say to the students?
You say, ‘____________________________’
14.3.2 Data analysis

Following the CCSARP coding scheme for the analysis of apology, the
participants’ apology realizations were classified into the following five

major strategies. One apology response may contain more than one
strategy.
(i) Direct apology. The speaker expresses an explicit apology with formu-
laic, routinized expressions such as ‘I’m sorry’, ‘Excuse me’.
(ii) Taking on responsibility. The speaker admits the offence, including self-
blame (‘It’s all my fault’), lack of intent (‘I accidentally left the book in
my car’), admission of fact (‘I haven’t corrected the exam papers yet’).
(iii) Explanation. The speaker gives an explanation or account of the situa-
tion in order to mitigate his/her guilt (‘I got an unexpected phone call
and lost track of time’).
(iv) Offer of repair. The speaker offers to remedy damage inflicted on the
offended party by specific compensation for the offence (‘I’ll return
the book as soon as possible’).
(v) Verbal redress. The speaker expresses concern for the offended party
(‘Are you OK?’), or promises of forbearance (‘I promise it won’t happen
again’).
14.3.3 Results and discussion

English and Chinese native controls’ apologizing behaviour
Comparisons of English and Chinese native controls’ apology realizations
indicated that both native control groups employed a similar amount of
apology strategies in each response (ENSs, 1.89; CNSs, 2.16), and they did
not differ significantly in the frequency usages of individual apology
strategies (Table 14.9). In addition, they had the same preference order
for the selection of apology strategies; they both adopted direct apologies
and taking on responsibility much more often than the other strategies.
The apology realizations of ENSs and CNSs appeared to be fairly similar.
A closer examination, however, revealed that the two control groups
varied in the overall apology realization patterns in relation to the social
factor of status. As we can see in Figure 14.4, ENSs employed a similar
amount of apology strategies across status conditions, but CNSs adopted
more strategies as the addressee’s status increased, and one-way ANOVA
Table 14.9 Distribution of apology strategies in the L1 control

groups (percentages) (from Su 2012, p. 246)
ENSs CNSs t-test p
Direct apology 44 37 1.94 n.s.

Taking on responsibility 27 29 −1.49 n.s.
Offer of repair 13 15 −1.16 n.s.
Explanation 10 9 0.37 n.s.
Verbal redress 4 6 −1.71 n.s.
Other 2 4 — —

312 I-RU SU
Table 14.10 Mean number of apology strategies in relation to social

variables in the L1 control groups (from Su 2012, p. 247)
ENSs CNSs t-test

Mean SD Mean SD t p
Status
High 3.70 1.37 3.20 1.30 1.45 n.s.
Equal 3.97 1.19 4.03 1.35 −0.20 n.s.
Low 3.60 0.97 4.67 1.06 −4.07 .000
ANOVA F = 0.77 n.s. F = 10.51, p = 0.000
Distance
Acquaintance 5.77 1.61 6.07 1.55 −0.73 n.s.
Stranger 5.47 1.76 5.80 1.58 −0.77 n.s.
t-test t = 0.83 n.s. t = 0.79 n.s.
Note. The maximum number in each category of status is 10 (5 major strategies × 2

situations); the maximum number in each category of distance is 15 (5 major
strategies × 3 situations).
6
Mean number of strategies
4
High
3
Equal
Low
2
0
ENSs CNSs
Figure 14.4 Mean number of apology strategies across status conditions in the L1 control
groups
analyses confirmed this observation (Table 14.10). Moreover, CNSs used

significant more strategies than ENSs when the offended party was of
higher status (t = −4.07, p = 0.000). These findings suggest that Chinese
native speakers were more status-sensitive in their apologizing behaviour
than their English counterparts, which corroborates the findings of pre-
vious studies on Chinese and English apologies (Barnlund and Yoshioka
1990; Tsai 2002).
English and Chinese native controls’ apology realizations were further
scrutinized by examining the frequency of each major apology strategy in
response to the social variables of status and distance. Results of statistical
analyses indicated that while ENSs adopted explicit apologies to a similar
extent regardless of the status of the offended party (F = 2.13, n.s.), CNSs
used significantly fewer direct apologies when apologizing to an addressee

of lower status than to one of equal or higher status (F = 21.56, p = 0.00).

Results of t-test analyses further showed that CNSs adopted significantly
fewer explicit apologies than ENSs when apologizing to an inferior
(t = 3.33, p = 0.002). Moreover, CNSs acknowledged responsibility (t = −3.05,
p = 0.003) and expressed verbal redress (t = −3.69, p = 0.000) significantly
more often than ENSs did when the offended party had higher power.
With respect to the factor of social distance, CNSs were more prone to
acknowledge responsibility to strangers (t = −2.09, p = 0.04) and to offer
more repairs (t = −2.48, p = 0.02) and verbal redress (t = −2.69, p = 0.01) to
acquaintances than ENSs.
Chinese EFL learners’ apologizing behaviour in their L2

Comparing the apology responses of Chinese EFL learners to those of ENSs,
we found that the three groups of participants provided a similar amount
of strategies in each response (ENSs, 1.89; intermediate EFL, 1.77; advanced
EFL, 2.07). They also shared similar frequency usages of individual apology
strategies, except for providing explanations. The intermediate learners
provided significantly fewer explanations to mitigate their offences than
ENSs and the advanced learners (ENSs, 10%; intermediate EFL, 4%;
advanced EFL, 10%). Nonetheless, the intermediate learners’ under-use of
providing explanations could not be attributed to L1 influence because the
apology realizations of English and Chinese native controls did not differ
in this regard. A probable explanation is that the intermediate learners did
not have sufficient linguistic facility to explain adequately for an offensive
act. A similar observation was reported in Trosborg’s (1987) study of
Danish learners of English.
Chinese EFL learners’ overall strategy use in relation to the social factors
of power and distance was further examined and compared to that of ENSs.
As shown in Table 14.11, the intermediate learners adopted significantly
Table 14.11 Mean number of apology strategies in relation to social vari-

ables in the English DCT (from Su 2012, p. 250)

Status
High 3.70 1.37 3.10 1.56 3.73 1.17 2.01 n.s.
Equal 3.97 1.19 3.70 0.95 4.03 1.27 0.71 n.s.
Low 3.60 0.97 3.83 1.05 4.17 0.87 2.59 n.s.
ANOVA F = 0.77 n.s. F = 3.08, p = 0.05 F = 1.18 n.s.
Distance
Acquaintance 5.77 1.61 5.60 1.63 6.23 1.45 1.32 n.s.
Stranger 5.47a 1.76 5.00b 1.93 6.17a 1.05 3.92 0.02
t-test t = 0.84 n.s. t = 1.89 n.s. t = 0.25 n.s.
Note. The maximum number in each category of status is 10, and the maximum number
in each category of distance is 15. Means with different alphabet letters within the same
category differ significantly (p < 0.05) by the post hoc Tukey test.

314 I-RU SU
fewer apology strategies when apologizing to inferiors than to equals and

superiors (F = 3.08, p = 0.05), whereas both ENSs and the advanced learners
employed a similar amount of strategies across status conditions; none of
the groups reacted to the factor of distance. The fact that the intermediate
group made their apology responses contingent on the direction of the
status relationship could be a result of L1 transfer, as CNSs also exhibited
the similar apologizing behaviour. By contrast, the advanced learners
seemed to have picked up the egalitarian target usage in this respect in
their L2 apologies.
With regard to the frequency of each apology strategy in response to
status and distance, overall there was little difference among the three
groups. What is noteworthy, however, is that while ENSs’ use of direct
apologies did not vary according to status relations (F = 2.13, n.s.), the
intermediate (F = 4.10, p = 0.02) and advanced (F = 10.41, p = 0.00) learners
did respond to the status factor; more specifically, both groups of learners
appeared to use significantly fewer direct apologies with people of lower
status (intermediate EFL, 70%; advanced EFL, 65%) than with people of
higher status (intermediate EFL, 89%; advanced EFL, 79%). This is an indica-
tion of L1 transfer as the Chinese native controls have been shown to be
more status-sensitive than their English counterparts in the use of direct
apologies.
Moreover, the advanced EFL learners were more prone to acknowledge
responsibility to strangers (p = 0.02) and offer more repairs to acquain-
tances (p = 0.001) than were the ENSs. This is also an indication of L1
influence, as CNSs also diverged from ENSs in these aspects. It seems
that the advanced learners experienced more negative transfer from the
L1 than the intermediate learners, contrary to most people’s expectations.
Previous research on interlanguage apology has noted that advanced lear-
ners are more capable of positively transferring their L1 strategies to L2
apologizing than lower-level learners because they have better linguistic
ability (Maeshiba et al. 1996; Shih 2006). It is probably fair to say that
advanced learners are also more likely than less proficient learners to
negatively transfer their L1 apologizing behaviour to their L2 because
they have better linguistic facility and can translate apology strategies
from the L1 to the L2.
Chinese EFL learners’ apologizing behaviour in their L1

Chinese EFL learners’ apology realizations were compared to those of CNSs
to see if the EFL learners carried the L2-specific strategies over to L1
apologizing. In particular, we would like to see if the Chinese EFL learners’
overall strategy use in relation to the social variable of status differed from
that of CNSs. Results of one-way ANOVA analyses indicated that the social
factor of status affected the apology responses of CNSs (F = 10.5, p = 0.000)
and those of the intermediate learners (F = 3.51, p = 0.03), but not the
advanced learners’ (F = 0.99, n.s.) (Table 14.12).

Table 14.12 Mean number of apology strategies in relation to social

variables in the Chinese DCT (from Su 2012, p. 253)

Status
High 3.20 1.30 3.50 1.70 3.87 1.04 1.78 n.s.
Equal 4.03 1.35 4.20 1.21 4.23 1.28 0.21 n.s.
Low 4.67 1.06 4.37 1.03 4.23 1.17 1.25 n.s.
ANOVA F = 10.5, p = 0.00 F = 3.51, p = 0.03 F = 0.99 n.s.
Distance
Acquaintance 6.07 1.55 5.97 1.54 6.00 1.74 0.03 n.s.
Stranger 5.80 1.58 6.17 1.80 6.37 1.69 0.86 n.s.
t-test t = 0.79 n.s. t = −0.69 n.s. t = −1.00 n.s.
Note. The maximum number in each category of status is 10; the maximum number in
each category of distance is 15.
5
Mean number of strategies
4
High
3
Equal
Low
2
0
CNSs Inter EFL Adv EFL
Figure 14.5 Mean number of apology strategies across status conditions in CNSs and EFL
learners
Figure 14.5 presents the mean number of apology strategies across

status conditions among CNSs and the EFL learners. As shown, CNSs and
the intermediate learners adopted considerably fewer strategies when
apologizing to inferiors than to equals and superiors, whereas the
advanced learners employed a similar number of strategies regardless of
the interlocutor’s relative status. For the advanced learners, this can be
an indication of L2 transfer as it has been shown that the variable of
social status did not affect ENSs’ overall strategy use but it did for that of
the CNSs.
Within-group comparisons in English versus Chinese

Paired-samples t-tests were conducted to see if the Chinese EFL learners
separated or merged their two apologizing repertoires. Results showed
that the intermediate learners seemed to behave differently in their two

316 I-RU SU
Table 14.13 Intermediate EFL learners’ mean number of

strategies in relation to social variables in English and Chinese
apologies (from Su 2012, p. 254)

Mean SD Mean SD t p
Status
High 3.10 1.56 3.50 1.70 −1.51 n.s.
Equal 3.70 0.95 4.20 1.21 −2.24 0.03
Low 3.83 1.05 4.37 1.03 −2.15 0.04
ANOVA F = 3.08, p = 0.05 F = 3.51, p = 0.03
Distance
Acquaintance 5.60 1.63 5.97 1.54 −1.41 n.s.
Stranger 5.00 1.93 6.17 1.80 −3.34 0.002
t-test t = 1.89 n.s. t = −0.69 n.s.
Note. The maximum number in each category of status is 10; the max-
imum number in each category of distance is 15.
languages when apologizing. They employed significantly more strategies

when redressing offenses in Chinese (2.02) than in English (1.77). They also
provided significantly more explanations in Chinese apologies than in
English apologies (8% versus 4%). In addition, as seen in Table 14.13, the
intermediate learners used significantly more strategies when apologizing
in Chinese than in English in half of the social conditions. Overall, it seems
that the intermediate learners had two separate apology systems.
However, this should not be the case as the apology realizations of ENSs
and CNSs did not vary in the aforementioned aspects. A possible explana-
tion for the variation in the intermediate learners’ apology responses in
their L1 and L2 is that their linguistic command was not great enough to
elaborate on their apology realizations in the L2, and so their apology
realizations were shorter in English than in Chinese.
The advanced learners, by contrast, appeared to have merged their two
apology systems. They made use of a similar amount of apology strategies
in English (2.07) and Chinese (2.06), and they had a similar frequency of use
of individual strategies in English and Chinese apologies. Moreover, the
results of statistical analyses indicated that there was no significant varia-
tion between their two languages in the overall strategy use in most of the
social situations (Table 14.14). The advanced learners appeared to apolo-
gize in essentially the same way in their two languages.
14.4 Conclusions
The two studies reviewed here addressed the issue of bi-directional trans-
fer in adult foreign language learners’ pragmatic competences from a
multi-competence perspective. Su (2010) investigated Chinese EFL

Table 14.14 Advanced EFL learners’ mean number of

strategies in relation to social variables in English and Chinese
apologies (from Su 2012, p. 255)

Mean SD Mean SD t p
Status
High 3.73 1.17 3.87 1.04 −0.45 n.s.
Equal 4.03 1.27 4.23 1.28 −0.70 n.s.
Low 4.17 0.87 4.23 1.17 −0.32 n.s.
ANOVA F = 1.18 n.s. F = 0.99 n.s.
Distance
Acquaintance 5.60 1.63 5.97 1.54 −1.41 n.s.
Stranger 5.00 1.93 6.17 1.80 −3.34 0.002
t-test t = 0.25 n.s. t = −1.00 n.s.
Note. The maximum number in each category of status is 10; and the
maximum number in each category of distance is 15.
learners’ requesting behaviour in their two languages. Transfer from both

directions (i.e. from L1 to L2 and from L2 to L1) was found to occur
simultaneously in Chinese EFL learners’ requesting behaviour.
Regardless of proficiency level, Chinese EFL learners were more direct
than English native controls when making requests in English, but at the
same time they were more indirect than Chinese native controls in for-
mulating Chinese requests. Moreover, both intermediate and advanced
learners differentiated their two requesting repertories in terms of strat-
egy use; they adopted significantly more conventionally indirect strategies
in formulating requests in English than in Chinese. The advanced learners
also employed more external modifiers in English requests than in
Chinese ones. These results suggest that foreign language learners may
have two separate pragmatic systems, which nonetheless may differ in
some way from that of native speakers of either language.
The finding that the Chinese EFL learners, especially the advanced
learners, exhibited a differentiated pattern of requesting behaviour in
their two languages is in contrast to previous research which evidenced
merged pragmatic systems in immigrants (Blum-Kulka 1990; Valdes and
Pino 1981). Compared to these immigrants, the EFL learners of the study
have rather limited exposure to English. These findings together suggest
that the language learning/use environment may play a role in how L2
users’ two language systems are integrated. The native language environ-
ment may facilitate the integration of the two languages of L2 users better
than the non-native environment. However, if L2 exposure is intensive,
convergence of pragmatic competences is also possible in foreign lan-
guage learners (Cenoz 2003).
Bi-directional transfer was also observed in Chinese EFL learners’
apology performance (Su 2012). The intermediate learners were more

318 I-RU SU
status-sensitive than the English native controls when apologizing in

English; the advanced learners were more prone to acknowledge respon-
sibility to strangers and offered more repairs to acquaintances than the
English native controls. L2 influence on the L1, on the other hand, was less
noticeable and only observed in the advanced learners. The advanced
learners did not vary their apology realization patterns according to the
interlocutor’s relative power, as Chinese native controls did when apolo-
gizing in the L1; they appeared to have applied the English egalitarian
usage to their L1 apology performance. As for whether the Chinese EFL
learners differentiated their two apologizing repertoires as they did with
the requesting systems, the results showed that the advanced learners
merged their two apologizing repertoires. The intermediate learners
appeared to perform differently in their two languages, but it is not so
much that they had separate apology systems as that their English profi-
ciency kept them from producing longer utterances in English, and so
their English apology responses were significantly shorter than the
Chinese ones.
It is interesting to note that the advanced Chinese EFL learners differ-
entiated their two pragmatic systems in terms of requesting repertoires
but they merged the apology ones. One possible reason is that Chinese and
English share more similarities in apology realizations than in request
realizations. Indeed, Chinese and English apologies produced by the native
speakers have been shown to share a similar preference order for the
selection of strategies and displayed similar frequencies of individual
strategies. On the other hand, in terms of request performance, the
Chinese and English native speakers differed from each other in the
strategy use; the English native speakers employed significantly more
conventionally indirect strategies than their Chinese counterparts.
English conventional indirectives such as ‘Could you . . .?’, ‘Can you . . .?’
are likely to be treated as formulas and explicitly taught in the foreign
language classroom, and this may be the reason for the significant differ-
ence in the frequency usage of this strategy in the EFL learner’s request
realizations in their two languages. In light of the two pragmatic studies
reviewed here, we can conclude that the patterns of bi-directional interac-
tion between foreign language learners’ pragmatic systems may vary
across speech acts; the integration or separation of the two pragmatic
repertoires may depend on the degree of similarity in the realizations of
a given speech act shared by the learner’s two languages and the way in
which a given speech act is learned.
The two studies reviewed here are among the few early attempts to
probe L2 users’ communicative competence from a multi-competence
perspective. The findings suggest that sociocultural rules for appropriate
speaking are also susceptible to bi-directional influence, even for foreign
language learners who have relatively limited exposure to the target lan-
guage and culture. More studies looking at different speech acts are

recommended to further our understanding of bi-directional interaction

of L2 users’ pragmatic competences.
References
Barnlund, D. C. and Yoshioka, M. 1990. ‘Apologies: Japanese and American

styles’, International Journal of Intercultural Relations 14, 193–206.
Bergman, M. L. and Kasper, G. 1993. ‘Perception and performance in native
and non-native apology’, in G. Kasper and S. Blum-Kulka (eds),
Interlanguage Pragmatics, pp. 82–107. Oxford University Press.
politeness in family discourse’, Journal of Pragmatics 14, 259–289.
Blum-Kulka, S., House, J. and Kasper, G. (eds) 1989. Cross-cultural Pragmatics:
Requests and Apologies. Norwood, NJ: Ablex.
Blum-Kulka, S. and Olshtain, E. 1986. ‘Too many words: length of utterance
and pragmatic failure’, Studies in Second Language Acquisition 8: 47–61.
Brown, A. and Gullberg, M. 2011. ‘Bidirectional cross-linguistic influence
in event conceptualization? Expressions of Path among Japanese lear-
ners of English’, Bilingualism: Language and Cognition 14, 79–94.
Brown, P. and Levinson, S. C. 1987. Politeness: Some Universals in Language
Usage. Cambridge University Press.
Byon, A. 2004. ‘Sociopragmatic analysis of Korean requests: pedagogical
settings’, Journal of Pragmatics 36, 1673–1704.
Cenoz, J. 2003. ‘The Intercultural Style Hypothesis: L1 and L2 interaction in
requesting behaviour’, in V. J. Cook (ed.), pp. 62–80.
Cenoz, J. and Valencia, J. 1996. ‘Cross-cultural communication and inter-
language pragmatics: American vs. European requests’, Pragmatics and
Chen, H.-L. 2006. A Study of Chinese EFL Interlanguage Requests. Unpublished
MA thesis, National Sun Yat-sen University, Taiwan.
Cook, V. J. (ed.) 2003. Effects of the Second Language on the First. Clevedon:
Dussias, P. E. and Sagarra, N. 2007. ‘The effects of exposure on syntactic
Gass, S and Neu, J. (eds) 1996. Speech Acts across Cultures. Berlin: Mouton De
Gruyter.
Gass, S. and Selinker, L. 1992. Language Transfer in Language Learning.

320 I-RU SU
Hymes, D. 1964. ‘Directions in ethnolinguistic theory’, American

Anthropology 6, 6–56.
Laufer, B. 2003. ‘The influence of L2 on L1 collocational knowledge and on
L1 lexical diversity in free written expression’, in V. J. Cook (ed.),
pp. 19–31.
Lee-Wang, S.-M. 1994. ‘Imperatives in requests: direct or impolite: observa-
tions from Chinese’, Pragmatics 4, 491–515.
Maeshiba, N., Yoshinaga, N., Kasper, G. and Ross, S. 1996. ‘Transfer and
proficiency in interlanguage apologizing’, in S. Gass and J. Neu (eds),
pp. 155–187.
Mennen, I. 2004. ‘Bi-directional interference in the intonation of Dutch
speakers of Greek’, Journal of Phonetics 32, 543–563.
23, 190–214.
Selinker, L. 1969. ‘Language transfer’, General Linguistics 9 (2), 67–92.
Shih, H.-Y. 2006. An Interlanguage Study of the Speech Act of Apology Made by EFL
Learners in Taiwan. Unpublished master’s thesis, National Sun Yat-sen
University.
Stavans, A. 2003. ‘Bilinguals as narrators: a comparison of bilingual and
monolingual Hebrew and English narratives’, Narrative Inquiry 13,
151–191.
Su, I-R. 2001. ‘Transfer of sentence processing strategies: a comparison of
L2 learners of Chinese and English’, Applied Psycholinguistics 22, 83–112.
Su, I-R. 2010. ‘Transfer of pragmatic competences: a bi-directional perspec-
tive’, Modern Language Journal 94, 87–102.
Su, I-R. 2012. ‘Bi-directional transfer in Chinese EFL learners’ apologizing
behavior’, Concentric: Studies in Linguistics 38, 237–266.
Trosborg, A. 1987. ‘Apology strategies in natives/non-natives’, Journal of
Pragmatics 11, 147–167.
Tsai, P.-C. 2002. A Study of the Speech Act of Apology in Chinese. Unpublished
master’s thesis, National Tsing Hua University, Taiwan.
Valdes, G. and Pino, C. 1981. ‘Muy a tus ordenes: compliment responses
among Mexican–American bilinguals’, Language in Society 10, 53–72.
Yu, M.-C. 1999. ‘Universal and cultural-specific perspectives on variation in
the acquisition of pragmatic competence in a second language’,
Pragmatics 9, 281–312.
Yu, M.-C. 2004. ‘Interlinguistic variation and similarity in second language
speech act behavior’, Modern Language Journal 88, 102–119.
Zampini, M. L. and Green, K. P. 2001. ‘The voicing contrast in English and
Spanish: the relationship between perception and production’, in
J. L. Nicol (ed.), One Mind, Two Languages, pp. 23–48. Boston: Blackwell.
Zareva, A. 2010. Multi-competence and L2 users’ associative links: being
unlike native-like. International Journal of Applied Linguistics 20, 2–22.

15
Multi-competence
and endangered
language revitalization
Tracy Hirata-Edds and Lizette Peter
Throughout the world there exist enclaves of indigenous and minority

languages subordinated and subsumed by broader, more dominant com-
munity languages. In many, if not most, cases these languages hang in a
precarious balance of survival that depends, to a large extent, on conscious
efforts to transmit languages to new users and maintain sociolinguistic
domains in which the endangered languages can flourish. Viewing lan-
guage use through the multi-competence perspective of an “overall system
of a mind or a community that uses more than one language” (Cook,
Chapter 1, this volume, p. 3) is particularly relevant to such contact situa-
tions since the phenomena of language contact, language shift and lan-
guage revitalization (LR) can all be characterized by the multiple
competencies of the users of the languages at stake.
In this chapter, we consider how the premises inherent in a multi-
competence view of language use provide a more robust and productive
framework for those involved in the revival of endangered languages than
does a monolingual model of second language acquisition. We begin by
establishing some critical components of language shift and then explore
specific ways in which the process of reversing language shift values users’
multiple competences as it attempts to rebuff monolingualism in a major-
ity or colonizing language and strives towards the reinstatement of a
multilingual way of life.
15.1 Language shift
Language shift commonly refers to the phenomenon that occurs when a

language community changes from one language to another for habitual,
everyday use (de Vries 1992). In a contact situation, several options arise
for users of a less dominant language, the most likely of which is eventual

322 T R A C Y H I R ATA - E D D S A N D L I Z E T T E P E T E R
adoption of the dominant language in a normalized pattern of language

use, at least in formal situations. As language norms change and use of the
new language seeps into more and more sociolinguistic domains (Fishman
1991), language shift occurs from monolingualism in the less dominant
language to bilingualism in both languages and then, potentially, to mono-
lingualism in the dominant language.
Language shift among indigenous and minority autochthonous peoples
over centuries of colonization, depletion of traditional lands, genocide, and
pressure to assimilate has resulted in wholesale extinction of vast numbers
of languages, drastically diminishing linguistic diversity on a global scale.
Worldwide, estimates place the current number of languages between six
and seven thousand (Lewis, Simons and Fennig 2013; Moseley 2010), of
which 37 per cent or more are no longer being transmitted to the next
generation in a sustainable way (Simons and Lewis 2013). Although lan-
guage shift has occurred since the beginning of human contact, linguistic
diversity has waned at notably higher rates over the past 35 years, with the
speed of language loss unexpectedly rapid and accelerating (Harmon and
Loh 2010; Harrison 2007; Whalen and Simons 2012). The shift of languages
at risk of extinction tends to follow a pattern of decreasing use in a variety of
sociolinguistic domains, especially among younger speakers, until even-
tually the language is no longer used in homes, a harbinger of loss because
intergenerational transmission is crucial for language survival (Fishman
1991; Kornai 2013; Lewis and Simons 2010).
Ominous predictions portend that as many as 90 per cent of the world’s
languages will no longer be vibrant one hundred years from now (Krauss
1992, 1996, 2007; Mithun 1999; UNESCO 2003; Warner 2001). Anderson
(2011) notes 20 diverse language “hotspots” worldwide where languages
are most at risk. By continent, language loss estimates range from 317
dead/dying languages in Australia and New Zealand, to 204 in South
America, 163 in North America, 131 in South-Eastern Asia and 81 in
Melanesia (Simons and Lewis 2013). Language communities at risk include
aboriginal peoples in Papua New Guinea, Australia, and the Americas;
national minorities in Africa, Asia (e.g., Ainu in Japan), and islands of the
Pacific (e.g., Māori in New Zealand); and minority autochthonous peoples
(e.g., Basque, Irish, and Welsh in Europe).
Although the process of language shift may appear linear and systema-
tic, it is, in fact, quite variable as communities, families and individuals
succumb to or resist the pressure to shift language use in different ways
and at different rates. Multi-competence is thus an apt characterization
of language shift at both the micro-level of the individual and the macro-
level of society, in that it connotes a dynamic and complex system of
bilingualism and biculturalism where linguistic repertoires and cultural
epistemologies collide and compete. Language revitalization, as the
inverse of language shift, can likewise be examined through the lens of
multi-competence.

Multi-competence and endangered language revitalization 323
15.2 Multi-competence and language revitalization
Language revitalization (LR) has in recent decades become a global move-

ment aimed at reversing the shift toward dominant-language monolin-
gualism and reinstating sociolinguistic domains in which the minority
language once thrived. Although the target of LR is ideally for future
generations to be equally proficient in both the dominant and ancestral
languages, actual success in revitalization is not as simple as language use
alone. A multi-competence perspective, which advocates language use
according to the preferences of the language user and values not only
the individual user’s potentiality, but also the framing of that potentiality
within a given community (Cook, Chapter 1), makes it well suited for
application to situations of language shift and revitalization. However,
some issues typically of less or no concern for more commonly taught
world languages are often critical issues for those aspiring to bring a lan-
guage back from decline through revitalization. Potentially more significant
within the topic of multi-competence are considerations of identity and
ideology, documentation and community language building.
15.3 Multi-competence and language identity
UNESCO’s Ad Hoc Expert Group on Endangered Languages recognizes that

“The extinction of each language results in irrecoverable loss of unique
cultural, historical and ecological knowledge. Each language is a unique
expression of the human experience of the world” (2003, p. 2). When a
language dies, then, the distinct perspectives (e.g., tradition, values, cul-
tural identity, community heritage) reflected by language and the distinct
linguistic characteristics of language itself (e.g., sounds, vocabulary, gram-
mar, discourse rules) become irrecoverable losses, especially felt by people
who feel connections with such knowledge. In fact, much motivation
among endangered language communities to revitalize their ancestral
tongue has been spurred by a perhaps strong version of linguistic relativ-
ity. Many advocates of LR feel endangered languages are worth saving
because of the relationship between language and culture, world view
and ways of life; indeed, many would say that the language IS the culture.
Such sentiments are germane to Cook’s Premise 3 (Chapter 1), that multi-
competence affects the whole mind. For users of endangered languages,
who strive to reconcile their indigenous identity with the dominant, main-
stream culture, any amount of competence in the ancestral language
provides the possibility for a cognitive balance of the different worlds
they live in.
When studying a second language for academic purposes (e.g., school
requirements), there may be little thought given to beliefs about or atti-
tudes toward the language; however, there is little contestation that when

indigenous peoples are engaged in activities specifically designed to main-

tain or revitalize their languages and cultures, they connect to something
ancestral that goes beyond language and deep into their psyche and sense
of identity. Any knowledge of the language can therefore represent access
to this identity, whether through the recitation of a memorized prayer or
chant, the ability to read the Bible in a heritage language, or knowledge of
vocabulary associated with medicinal properties of certain plants and
animals. And yet, for many users of endangered languages, such a limited
range of functional ability is indicative of a lack of cultural competence.
The notion of “native speaker,” in this sense, has interesting connotations
in terms of indigenous identity. A proficient “non-native” speaker of
Mohawk (lower case “native,” implying first language speaker), for exam-
ple, may be viewed as less “Native” (upper case “Native,” implying indi-
genous) than a “native” speaker of Mohawk. In Cook’s (this volume) sense,
“native” pertains to national identity; in the indigenous sense, native may
relate more to a Native identity.
Thus, language competence in an indigenous tongue is often associated
with cultural competence, which is indelibly tied to identity. Some elders,
teachers, or even learners of an endangered language may have emotional
connections to the language and/or be critical of skills that veer from some
idealized “standard,” resulting in a less than supportive environment for
LR to be successful. Self-consciousness about potentially making mistakes,
being made fun of, or being disapproved of affects users no matter their
skill level, including elder native speakers (Hinton and Hale 2001). Peter
(2014) notes that Cherokee learners in Oklahoma were sometimes hesitant
to converse with more fluent speakers not only out of fear of criticism, but
also of being perceived as less “Cherokee.” She found prevalent an “ideol-
ogy of legitimacy” in which language proficiency becomes a social con-
struct in and of itself, with new ideals of “expertise” resulting in a
redistribution of status and legitimacy within the community. Jaffe
(1999) further explains that “language shift also reworks the value of
‘old’ communicative competencies: no longer taken-for-granted resources
of most adult speakers, they become heightened indices of cultural
authenticity and social belonging” (p. 2).
The relationship between language and culture may be especially bewil-
dering for indigenous youth, who struggle to make sense of their own
linguistic multi-competences. McCarty, Romero-Little and Zepeda (2008),
for instance, find that Navajo youth in the southwestern USA perceived
their heritage language as a part of who they were and viewed bilingualism
as necessary for functioning in both the Navajo and non-Navajo world. Yet,
contradictorily, many also saw their Navajo ability as a source of embar-
rassment, to the extent that even proficient speakers among them were
reticent about using the language. The result of this stigmatization is not
only that they missed natural and culture-rich opportunities for interac-
tion with adult speakers, but also that adults in the community were often

unaware that the youth were even able to speak Navajo at all. If language is
important for maintaining indigenous identity in a world that socially and
politically seeks to diminish it, then special care is needed to impart
indigenous knowledge to the youth (Hinton 2013b) and to instill an appre-
ciation that diminishes reticence and enhances dignity in use of the
language. From a multi-competence perspective, this also means that
second language users of ancestral languages are entitled to be considered
for what they are capable of, not for lacking in comparison to a native
speaker standard (Cook 2003).
For younger generations, in particular, technology has opened up seem-
ingly endless possibilities for translingual practice, in which texts and talk
“are meshed and mediated by diverse codes” and communication
resources “gain form in situated contexts for specific interlocutors in
their social practice” (Canagarajah 2013, pp. 6–7). Television programming
in indigenous languages, for example, has succeeded in both exposing
viewers to those languages and demonstrating their right to occupy the
airwaves. The advent of the internet and mobile technology has increased
even more significantly the opportunities for language use among indivi-
duals with multiple language competences. Consider the case of youth and
young adult users of the Cherokee language, a severely endangered lan-
guage spoken primarily in the US states of North Carolina and Oklahoma.
With the creation of a global Unicode font in the late 1990s for the
Cherokee writing system known as Syllabary, the Cherokee language
was quickly propelled into cyberspace. Tehee (2014) lists a number of
operating systems that have been translated into Cherokee (e.g., Apple
Mac, iPhone, and iPad; Windows 9 and its online Office suite of Word,
PowerPoint, Excel, and OneNote) as well as the creation of Cherokee
versions of Wikipedia, Facebook, and Twitter. As Tehee notes, the digitiza-
tion of Cherokee is driven primarily by L2 users, some of whom refer to
themselves as “Tsalageeks” (Tsala stemming from Tsalagi, the Cherokee
word for “Cherokee”) and who have co-opted this digital domain as a space
for second language users to communicate in and actually create language,
thereby exercising their right to speakerhood without fear of scorn for
their mistakes or perceived limitations.
15.4 Multi-competence and language documentation
Language documentation has traditionally involved linguists working

with “native” speakers of a language in the field to record on paper or
electronically the intricacies of a language’s components, including gram-
mar, phonology, syntax, lexicon, and pragmatics. The outcome of this
work served primarily academic purposes: the compilation of data from
which to extract meaning, make connections and seek to comprehend
human linguistic capabilities. The disappearance of endangered languages

and language families results in the loss of evidence about the range of
possibilities in language, a devastating blow for linguistic theory because
of the phenomena that would be rendered undiscoverable (Whalen and
Simons 2012). Testing and building theories about language universals,
mental processes and learning processes, for example, are not scientifi-
cally viable without information from all languages within a family.
Though academically interesting, linguistic research can be of limited
interest to a community focused on LR because data for scientific inquiry
are often not ideally suited for revitalization. Indeed, it is not uncommon
for indigenous community members to view the linguistic process of
documentation as only being for the sake of storage or preservation, rather
than contributing to a language’s vitality and growth (Hinton 2001b).
Documentation, whether in the form of word lists, dictionaries, and
grammars, or audio/visual recordings, cannot alone save a language; it
does, however, comprise a critical facet of the revitalization enterprise.
For the most severely endangered languages (i.e., only elderly speakers, no
children or young adult speakers) or “sleeping” languages (i.e., no living
fluent speakers), documentation forms the foundation for language study,
and a number of endangered language communities have successfully
used documentation to bring a language back into existence, with
Hebrew, which was not a language of the home for two thousand or so
years, being one of the most exceptional cases to date (Hinton 2001c). For
documentation to be constructive for LR, Penfield and Tucker (2011) sug-
gest that among many areas of need are applied linguistic training and
team approaches that result in materials useful to linguists and accessible
to the endangered language community. Furthermore, Yamada (2007,
2011) advocates partnerships that meet both academic and speech com-
munity needs.
In addition to the form and usability of documentary data, Cook
(Chapter 1, this volume) raises the issue of whether or not judgments
made by multilingual informants can be considered accurate for the sake
of documentation under traditional research conditions, “by virtue of the
influence of the second language that each of these linguists knows”
(p. 10). Given the historical trajectory of language shift and the multi-
competent users left in its wake, however, monolingual speakers of most
endangered languages are rare, if not non-existent, and have been so for a
century or more. Moreover, it is the exception in many language “hotspots”
to find a speaker of an endangered indigenous language who acquired the
language as an infant and has spoken it continuously throughout life.
Therefore, no grammar or dictionary of an endangered language created
from elicited data is likely to encompass a “complete” language; rather, it
should be considered representational of the language as used by that
person just as its users characterize a multi-competent speech community.
Premise 1 of multi-competence – that it concerns the total language
system in a single mind or community – provides a useful way to understand

LR documentation practices. From a monolingual viewpoint, the ideal

native speaker is not only monolingual, but also speaks the status version
of the language rather than a “non-standard” variety. A multi-competence
view, however, acknowledges that in most endangered language contexts
where a “standard” version of the language has never existed, all users
speak some kind of idiolect, and so-called “native” speakers themselves
may have varying levels of proficiency. It is quite possible that different
dialects spoken among endangered language communities actually reflect
different families’ experiences with language shift, as use of the mother
tongue may have been interrupted by schooling, periods of distance from
the family, exposure to written text in the language, and the like. In terms of
documentation, then, multi-competence brings to the forefront, indeed
values, variability of language use among community members rather
than imposing an arbitrary “standard.”
15.5 Multi-competence and community language

building
As noted earlier, languages disappear for myriad combinations of factors

often related to contact between language groups of differing power and
status, with results favoring some languages and marginalizing others
economically, environmentally, politically and culturally (Crystal 2013,
2000; Hinton 2013b; McCarty, Romero-Little and Zepeda 2008). One way
this kind of repression is manifested is in societal and linguistic change
(Aitchison 1991). That is, if a language is not used for key components of
life (e.g., government, education, commerce, communication), its speak-
ers are at risk of seizure of lands, monolingual language policies, and even
genocide, all of which shift communication to majority, dominant lan-
guages in ever increasing sociolinguistic domains (McCarty, Romero-Little
and Zepeda 2008).
The prospect for language revitalization is not a return to some idealized
“traditional” period. Few engaged in LR have any illusion of or desire for a
population of monolingual speakers of a once-endangered language. The
reality is that future generations of revitalized language users will speak a
modern version that will likely look markedly different from that of their
ancestors. So, the goal of LR is not only for younger generations to be able
to speak to elders – if elder users of the language even exist; it is also to
promote a new generation of users who will innovate the language to suit
their sociolinguistic needs, much in the way of modern Hebrew (Hinton
2001c). But even this more realistic scenario requires the reinstatement of
the language in traditional domains, including homes, ceremonial
grounds, and meeting houses, as well as in new domains traditionally
devoid of the language, such as schools, universities and professional
enterprises. A multi-competence perspective on community language

building supports the potential that all users of the endangered language
will contribute to its survival in any of these domains.
Communities dedicated to LR are most successful when they adopt a
realistic outlook of the language’s potential viability in contemporary
settings. Homes and families are the customary place for language trans-
mission, and infusing daily life with an endangered language serves to
incorporate the naturalness of language learning and use. This can be
done in situations where parents and/or elder family members are still
speakers of a language, as in Karuk, a Native American language isolate
spoken in northwestern California (Albers and Albers 2013); where par-
ents are second language users, as in Hawaiian (Wilson and Kamanā
2013) and Māori (O’Regan 2013); and even with “sleeping” languages
when there are no speakers, and parents must learn the language them-
selves using linguistic documentation alone, such as Miami-Illinois
(Baldwin et al. 2013), a Native American Algonquian language, histori-
cally of the southern Great Lakes region, and Wampanoag, a Native
American Algonquian language historically spoken in New England
(baird 2013). In all of these cases, families seek to raise children to be
multicultural and multi-competent in their ancestral culture and lan-
guage in addition to that of the pervasive environment. The surrounding
majority language has its place, certainly, but families seek to provide
nurturing for the endangered language to impart to children their “birth-
right to their language” (baird 2013, p. 23) through time and opportu-
nities for use. For these children, the human potentiality to know and use
more than one language (Cook, Chapter 1, this volume) is the norm – is
their norm.
Taking the language-in-the-home model a step further, some commu-
nities establish physical enclaves of language users where the majority
language may be present, but is not necessarily promoted in the way that
the target endangered language is. People who gather intentionally in
physical proximity share an understanding that only through an environ-
ment rich with potential for the language to flourish is there any hope of
interrupting the trend toward language loss. Irish speakers living in
Belfast, Northern Ireland, within a community of like-minded families,
for example, benefit not only from being able to use the language among
family members for daily life, but also from having a neighborhood Irish
support system in a region that is primarily English-speaking (Mac Póilin
2013). Of note in terms of multi-competence are the intergenerational
differences in language use and abilities: parents may be L2 users of
Irish raising children as L1 Irish speakers, who then raise their children
as L1 Irish speakers. Ideally, by the third generation, a mixed-age, multi-
competent population has the flexibility in their linguistic skills to ably
read, write, listen to, and speak Irish and English in a range of socio-
linguistic domains of equal status. In this way, language revitalization
aspires to transcend a diglossic condition in which one language is used

for purposes of wider communication, while the endangered language is

limited primarily to use in homes and neighborhoods; the goal is to
actually situate the endangered language “into the upper reaches of socio-
symbolic life” (Fishman 2006, p. 97).
An especially popular and potentially successful way to infuse a com-
munity with occasions to see and hear a vital language around them is
immersion schooling. For endangered language revitalization, this
approach introduces the language in a domain where it was tradition-
ally discouraged – schools – and dedicates the endangered language as
the medium of instruction to promote multi-competence through a
range of academic and social language experiences. The immersion
approach has been adopted by a number of endangered indigenous
language groups, including the Hawaiian, Māori, Acoma, Arapahoe,
Blackfeet, Cree, Cherokee, Hualapai, Mohawk, Mississippi Band of
Choctaw, and Navajo (Hinton 2001a; Hinton and Hale 2001; Lambert,
Genesee, Holobow and McGilly 1984; McCarty and Watahomigie 1999;
Peter and Hirata-Edds 2006; Stiles 1997; Wilson and Kamanā 2001,
2011), as well as by autochthonous groups such as Irish, Welsh,
Basque and Frisian in Europe (Gorter 2015). Although the focus of
immersion is on the endangered language, the goal is essentially one
of multi-competence: for students to be able to communicate fluently in
the majority language, which they are widely exposed to as soon as they
leave the linguistic confines of immersion, and to achieve solid commu-
nicative footing in the endangered language, usually coupled with lit-
eracy in both languages, as well.
Multi-competence beyond bilingualism is promoted in the Hawaiian
Nawahi program, with English introduced in 5th grade (through the med-
ium of Hawaiian) and additional heritage and ancestral languages added
later, including Latin in 7th and 8th grades, Japanese in 1st to 6th grades,
and Chinese in 7th to 12th grades (Wilson and Kamanā 2011). By inten-
tionally creating an atmosphere in which language variety is customary
and ever-present, children come to accept multilingualism as status quo in
the sense that Cook (Chapter 1) views the normalcy of knowing more than
one language. So, whether intentional or not, immersion programs foster
multi-competence of mind and community use of more than one lan-
guage (Cook, Chapter 1). Peter and Hirata-Edds (2009), for instance, note
how the role of literacy in the Cherokee Syllabary has been transformed
by LR efforts and especially by the near-exclusive use of the Syllabary’s
phonetic representations of syllables in an immersion school (rather
than a Romanized phonetic system that also exists for Cherokee). They
discovered that, for most of the children in the immersion school,
Cherokee is the first language they learn to read and write, even though
from an oral/aural standpoint Cherokee is their second language. By
contrast, their native Cherokee-speaking immersion teachers reported
that, even when Cherokee was the language spoken in their homes, for

each of them, English was the language they first learned to read in and
all but one indicated that they did not learn to read Cherokee Syllabary
until they were adults. The teachers’ and students’ bilingualism and
biliteracy occur on opposing ends of an oral–literate and monolingual–
bilingual continuum (Hornberger 2003), and these children were stron-
ger decoders (i.e., able to sound out/pronounce the sounds) of Cherokee
Syllabary than many of their teachers, even while their oral proficiency
was still developing.
For prospective endangered language users who may not have access to
learning in either their homes or a school environment, a specialized
approach for adult learners, known as master–apprentice, pairs an elder
speaker and a younger learner/user who spend time together doing cultu-
rally and contextually rich activities (e.g., discussions, story-telling, cook-
ing, building, exploring nature, playing games) exclusively in the target
language (Hinton 1994; Hinton and Hale 2001; Hinton, Vera and Steele
2002). The emphasis on authentic, functional communication adheres
nicely to Cook’s (Chapter 1, this volume) emphasis on communicative
use and ability rather than on linguistic limitations. Master–apprentice
as an LR model originated in California for American Indian languages
(Hinton 1994; Hinton and Hale 2001; Hinton, Vera and Steele 2002) and has
quickly been adopted throughout North America from First Nations’ com-
munities in British Columbia (Virtue, Gessner and Daniels 2012) to
Oklahoma Native communities, including Sauk (Sorensen and Weston
2011) and Chickasaw (Christopher 2010).
In addition to the obvious benefit of such intensive, hands-on language
learning sessions, master-apprenticeships create opportunities for
elders to speak their language that they otherwise would no longer
have and establish a direct intergenerational transmission mechanism,
thereby boosting future possibilities for apprentices to pass the language
on to others. Success in a master–apprentice program lies in the extent
to which the apprentice is able to capitalize on his or her individual
learning toward the benefit of the community. Whatever the apprentice
is able to teach to family and friends will depend not just on his or her
overall fluency, but equally on his or her ability to transmit the language
to others in functional and meaningful ways. In fact, according to Cook’s
(Chapter 1) Premise 2 of multi-competence, “Non-native speaker tea-
chers may make better role models for the students because they have
travelled the same route as them and are living exemplars of successful
L2 users able to handle two languages at the same time” (p. 14). In
master–apprentice as well as other endangered language situations, it
is often the case that L2 users possess greater metalinguistic awareness
of the language than their “master” counterparts, who may speak with a
high degree of proficiency yet possess little metalinguistic awareness. It
is in the combination of different degrees and kinds of competences that
LR efforts stand the greatest chance of success.

15.6 Measuring success of language revitalization efforts
From a multi-competence perspective (Cook, Chapter 1), a monolingual

native speaker is not an appropriate index by which to measure an L2
user’s abilities. This perspective is all the more important in language
contexts where there may not even be monolingual speakers with “native”
fluency and when the goal extends beyond mere language study. Although
attainment of proficiency is certainly desirable, success in revitalization is
also measured with respect to complementary outcomes. Some goals for
revitalization programs include documentation and linguistic knowl-
edge of the endangered language, and incorporation of cultural nuances,
as depicted through language, into daily life. Additionally, beyond this,
many programs also strive to mentor children, who will be the future for
their communities and who have a strong self-identity with language
iconic in that process.
If a monolingual view of success were applied to endangered language
revitalization, investigation likely would center on the similarities and
dissimilarities between L2 users of the language and those who speak the
language “natively.” There is value in research of this nature. In fact, a case
can be made that documenting the acquisition of language by L2 users
contributes to the documentation of the endangered language writ large,
so elicitations like grammaticality judgments and error analyses have
implications for both pedagogical practices and curriculum design within
more formal language learning settings, such as immersion schools (see
Peter and Hirata-Edds 2006; Peter, Hirata-Edds and Montgomery-Anderson
2008; Peter, Sly and Hirata-Edds 2011).
However, defining success solely on the basis of what an individual can
and cannot do in the language neglects the multiple ways in which that
individual interacts as part of a larger endangered language community.
And, since language revitalization is almost always a multifaceted enter-
prise with diverse and varying components that can be more or less
formal, more or less consistent and continuous and more or less inclusive
of a wide range of stakeholders, success is often measured in terms of
smaller undertakings rather than grand achievements. Language loss has
taken place over many centuries, with a concerted chipping away at
components of language and places for its use. The journey back to steady
and stable use will likewise take time, and the path must be differentially
appropriate for each situation. If one apprentice manages to get his or her
family to replace some English vocabulary with expressions learned
through interactions with an elder speaker, this is a version of success. If
one mother learns to sing an Irish lullaby to her newborn, this is another
kind of success.
Success can also be gauged in terms of the effect that the LR efforts have
on the community in terms of raised awareness of the status of the

language and improved attitudes toward its survival. An illustrative exam-

ple comes from Peter’s (2007) ethnography of the inaugural year of a
Cherokee immersion preschool. Although samples of language revealed
that children had limited functional abilities in Cherokee at the end of one
year, there were indications that awareness of these Cherokee-speaking
children within the broader community had sparked a renewed sense of
interest and pride in the language. That was a step forward. For Miami-
Illinois, a “sleeping” language, focus on creating speakers right now is not
necessarily a primary objective, but preparing for future revitalization
efforts is. Daryl Baldwin, Director of the Myaamia Center at Miami of
Ohio University, said that he is comforted by the knowledge that the
next generation “knows more about being Myaamia than I knew at that
age. Our language, in the context of our culture, provided that growth
experience” (Baldwin et al. 2013, p. 18). This understanding of self by the
younger generation is echoed in Hana O’Regan’s description of her chil-
dren’s “strong sense of cultural and personal identity that is informed by a
Māori worldview” (2013, p. 99) as opposed to her own childhood concerns
about legitimacy.
One goal of the Māori, to raise children as language users, required
active participation of not only parents and teachers, but also local elders
and other Māori community members. In addition, adding school as a new
domain for the language and serving as an inspiration to other grassroots
revitalization initiatives were important indicators of their success (King
2001). Defining success in similar yet broader measures comes from the
Hawaiian immersion program, which educates children in a system that
prioritizes the Hawaiian language while simultaneously raising them to be
vital assets to the community (Wilson and Kamanā 2011). Now that grad-
uates of the program are bringing up their children as speakers of
Hawaiian, the value of the language and its role in Hawaiian identity are
also being transmitted. Success comes in both the multi-competence of
young adults and in their valuing of this condition. These examples illus-
trate that, as members of different language minorities find their voice and
cherish their linguistic heritage and rights, young people will both make
broad impacts and daily personal choices “to develop meaningful roles for
their languages in their everyday lives and find good reasons to speak and
transmit them to their children” (Yamamoto, Brenzinger and Villalón
2008, p. 68), defining their view of success in LR.
15.7 Prospects for endangered languages
Endangerment is potentially in the future for many of the world’s lan-

guages, but recent interest in language revitalization offers hope that
current trends toward language shift might be interrupted for some indi-
viduals and communities. It is unclear what, precisely, revitalized

languages may look like. They may be hybrids or creolizations influenced

by both the L1 and the L2. This may result in changes to tonal systems, for
example, with flattening of tones or patterns more like that of the domi-
nant language. The morphology could lose some of its complexity or take
on characteristics related to both languages. The lexicon could expand
with words created to address a modern world or with words borrowed
from other sources. No one can say with any certainty what will happen
except that the manifestations will likely differ due to the distinct qualities
of each language community. The language spoken by graduates of the
Hawaiian immersion program, for example, does not sound quite the
same as it is spoken by “native” speakers, to the extent that those influ-
enced by a monolingual perspective may worry about “a corrosive force
that contaminates the purity and weakens the integrity of the [Hawaiian]
language” (Wong 1999, p. 95). But if we measure success in terms offered
by a multi-competence perspective, then the fact that several generations
of Hawaiians participate in a range of linguistic behaviors, translanguaging
between structures and modalities (Li Wei 2011) and within an ever-
expanding number of sociolinguistic domains, we can say that Hawaiian
language revitalization is a success story.
The question for endangered language communities may not be simply
whether or not the language can survive another generation, but rather
how to promote multi-competence in an era that favors homogenization
and standardization over diversity (McCarty 2003). Youth, in particular, are
pressured to conform to the norms of the dominant society and, although
very young minority language speakers may not have an awareness that
their language lacks status, as Gorter (2015) suggests, “as speakers they
soon become aware of the social inequalities of the languages they speak;
they pay attention to the significance of language choice and to the role of
languages in social relationships” (p. 5). If such awareness can silence a
young language user, then language revitalization efforts promoting the
advantages of multi-competence can counteract such reticence by instil-
ling confidence in and valuing use of ancestral languages together with
languages of the broader community.
References
Aitchison, J. 1991. Language Change: Progress or Decay? Cambridge University

Press.
Albers, P. and Albers, E. (Supahan) 2013. ‘Karuk language and the Albers
basket’, in L. Hinton (ed.), pp. 33–40.
Anderson, G. D. S. 2011. ‘Language hotspots: what (applied) linguists and
education should do about language endangerment in the twenty-first
century’, Language and Education 25, 273–289.
baird, jessie little doe 2013. ‘Wampanoag’, in L. Hinton (ed.), pp. 19–30.

Baldwin, D., Baldwin, K., Baldwin, J. and Baldwin, J. 2013. ‘myaamiaata-

weenki oowaah: “Miami spoken here” ’, in L. Hinton (ed.), pp. 3–18.
Canagarajah, S. 2013. Translingual Practice: Global Englishes and Cosmopolitan
Relations. London: Routledge.
Christopher, D. 2010. ‘Chickasaw culture in Oklahoma: Chickasaw medi-
cine, language and books, Chickasaw press, Chickasaw arts and the
Chickasaw cultural center’, NewsOK Online.
Cook, V. J. 2003. Effects of the Second Language on the First. Clevedon, UK:
Crystal, D. 2000. Language Death. Cambridge University Press.
Crystal, D. 2013. ‘The death of a language’, Languages Today, autumn, 14–15.
De Vries, J. 1992. ‘Language maintenance and shift: problems of measure-
ment’, in W. Fase, K. Jaspaert and S. Kroon (eds), Maintenance and Loss of
Minority Languages, pp. 211–222. Amsterdam: John Benjamins.
Fishman, J. 1991. Reversing Language Shift: Theoretical and Empirical Foundations
of Assistance to Threatened Languages. Clevedon, UK: Multilingual Matters.
Fishman, J. 2006. ‘What is reversing language shift (RLS) and how can it
succeed?’, in N. H. Hornberger and M. Pütz (eds), Language Loyalty, Language
Planning, and Language Revitalization: Recent Writing and Reflections from Joshua
A. Fishman, pp. 79–112. Bristol, UK: Multilingual Matters. Originally in
Journal of Multilingual and Multicultural Development (1990) 11 (1–2), 5–36.
Gorter, D. 2015. ‘Multilingual interaction and minority languages: profi-
ciency and language practices in education and society’, Language
Teaching 48 (1), 82–98.
Harmon, D. and Loh, J. 2010. ‘The index of linguistic diversity: a new
quantitative measure of trends in the status of the world’s languages’,
Language Documentation and Conservation 4, 97–151.
Harrison, K. D. 2007. When Languages Die: The Extinction of the World’s Languages
and the Erosion of Human Knowledge. New York, NY: Oxford University Press.
Hinton, L. 1994. Flutes of Fire: Essays on California Indian Languages. Berkeley,
CA: Heyday Books.
Hinton, L. 2001a. ‘Teaching methods’, in L. Hinton and K. Hale (eds),
pp. 179–189.
Hinton, L. 2001b. ‘New writing systems’, in L. Hinton and K. Hale (eds),
pp. 239–250.
Hinton, L. 2001c. ‘Sleeping languages: can they be awakened?’, in L. Hinton
and K. Hale (eds), pp. 413–417.
Hinton, L. (ed.) 2013a. Bringing Our Languages Home: Language Revitalization for
Families. Berkeley, CA: Heyday Books.
Hinton, L. 2013b. ‘Introduction’, in L. Hinton (ed.), pp. xi–xx.
Hinton, L. and Hale, K. 2001. The Green Book of Language Revitalization in
Practice. San Diego, CA: Academic Press.
Hinton, L., Vera, M. and Steele, N. 2002. How to Keep Your Language Alive: A
Commonsense Approach to One-on-One Language Learning. Berkeley, CA:
Heyday Books.

Hornberger, N. H. 2003. ‘Continua of biliteracy’, in N. H. Hornberger (ed.),

Continua of Biliteracy: An Ecological Framework for Education Policy, Research,
and Practice in Multilingual Settings, pp. 3–34. Bristol: Multilingual Matters.
Jaffe, A. 1999. Ideologies in Action: Language Politics in Corsica. New York, NY:
Mouton De Gruyter.
King, J. 2001. ‘Te kōhanga reo: Māori language revitalization’, in L. Hinton
and K. Hale (eds), pp. 119–131.
Kornai, A. 2013. ‘Digital language death’, PLoS ONE 8 (10): article e77056.
doi:10.1371/journal.pone.0077056
Krauss, M. 1992. ‘The world’s languages in crisis’, Language 68, 1–42.
Krauss, M. 1996. ‘Status of Native American language endangerment’, in
G. Cantoni (ed.), Stabilizing Indigenous Languages, pp. 16–21. Flagstaff, AZ:
Center for Excellence in Education.
Krauss, M. 2007. ‘Mass extinction and documentation: the race against
time’, in O. Miyaoka, O. Sakiyama and M. Krauss (eds), The Vanishing
Languages of the Pacific Rim, pp. 3–24. Oxford University Press.
Lambert, W. E., Genesee, F., Holobow, N. and McGilly, C. 1984. An Evaluation
of a Partial Mohawk Immersion Program in the Kahnawake Schools. Montreal:
McGill University.
Lewis, M. P. and Simons, G. F. 2010. ‘Assessing endangerment: expanding
Fishman’s GIDS’, Revue Roumaine de Linguistique 55, 103–120.
Lewis, M. P., Simons, G. F. and Fennig, C. D. 2013. Ethnologue: Languages of the
World, 17th edition. Dallas, TX: SIL International. www.ethnologue.com
of Pragmatics 43, 1222–1235.
Mac Póilin, A. 2013. ‘Belfast’s Neo-Gaeltacht’, in L. Hinton (ed.),
pp. 141–163.
McCarty, T. L. 2003. ‘Revitalizing Indigenous languages in homogenising
times’, Comparative Education 39, 147–163.
McCarty, T. L., Romero-Little, M. E. and Zepeda, O. 2008. ‘Indigenous
language policies in social practice: the case of Navajo’, in K. A. King,
N. Schilling-Estes, L. Fogle, J. J. Lou and B. Soukup (eds), Sustaining
Linguistic Diversity: Endangered and Minority Languages and Language
Varieties, pp. 159–172. Washington, DC: Georgetown University Press.
McCarty, T. and Watahomigie, L. J. 1999. ‘Preservation on the reservation
(and beyond): reclaiming Indigenous languages’, in Common Ground:
Archaeology and Ethnography in the Public Interest. National Park Service,
US Department of the Interior.
Mithun, M. 1999. The Languages of Native North America. Cambridge University
Press.
Moseley, C. (ed.) 2010. Atlas of the World’s Languages in Danger, third edition.
Paris: UNESCO Publishing. www.unesco.org/culture/en/endangeredlan
guages/atlas
O’Regan, H. 2013. ‘My language story’, in L. Hinton (ed.), pp. 80–100.

Penfield, S. and Tucker, B. V. 2011. ‘From documenting to revitalizing

endangered language: where do applied linguists fit?’ Language and
Education 25, 291–305.
Peter, L. 2007. ‘Our beloved Cherokee: a naturalistic study of Cherokee
preschool language immersion’, Anthropology and Education Quarterly 38,
323–342.
Peter, L. 2014. ‘Language ideologies and Cherokee revitalization: narra-
tives of impracticality, legitimacy, and hope’, Journal of Immersion and
Content-Based Language Education 2, 96–118.
Peter, L. and Hirata-Edds, T. 2006. ‘Using assessment to inform instruction
in Cherokee language revitalization’, International Journal of Bilingual
Education and Bilingualism 9, 643–658.
Peter, L. and Hirata-Edds, T. 2009. ‘Learning to read and write Cherokee:
toward a theory of literacy revitalization’, Bilingual Research Journal 2,
207–227.
Peter, L., Hirata-Edds, T. and Montgomery-Anderson, B. 2008. ‘Verb devel-
opment by children in the Cherokee language immersion program with
implications for teaching’, International Journal of Applied Linguistics 18,
166–187.
Peter, L., Sly, G. and Hirata-Edds, T. 2011. ‘Using language assessment to
inform instruction in Indigenous language immersion’, in D. J. Tedick,
D. Christian and T. Williams Fortune (eds), Immersion Education: Practices,
Policies, Possibilities, pp. 187–208. Bristol, UK: Multilingual Matters.
Simons, G. F. and Lewis, M. P. 2013. ‘The world’s languages in crisis: a
20-year update’, in E. Mihas, B. C. Perley, G. Rei-Doval and K. M. Wheatley
(eds), Responses to Language Endangerment: In Honour of Mickey Noonan; New
Directions in Language Documentation and Language Revitalization, pp. 3–19.
Amsterdam: John Benjamins.
Sorensen, B. and Weston, J. 2011. ‘I love Sauk language’, Cultural Survival 35.
http://issuu.com/culturalsurvival/docs/csq-35-4
Stiles, D. B. 1997. ‘Four successful Indigenous Language programs’, in
J. Reyhner (ed.), Teaching Indigenous Languages, pp. 148–262. Flagstaff,
AZ: Northern Arizona University.
Tehee, C. 2014. Negotiating Acceptance: A Sociocultural Analysis of Second
Language Users’ Construction of Speakerhood in Cherokee Nation Language
Revitalization. Unpublished doctoral dissertation, University of Oklahoma,
Norman, OK.
UNESCO 2003. Language Vitality and Endangerment. Ad Hoc Expert Group on
Endangered Languages: Programme Safeguarding of Endangered
Languages. Paris.
Virtue, H. A., Gessner, S. and Daniels, D. Xway’Waat 2012. B.C.’s Master–
Apprentice Language Program Handbook. British Columbia: First Peoples’
Cultural Council.
Warner, S. L. No’eau 2001. ‘The movement to revitalize Hawaiian language
and culture’, in L. Hinton and K. Hale (eds), pp. 133–146.

Whalen, D. H. and Simons, G. F. 2012. ‘Endangered Language Families’,

Language 88, 155–173.
Wilson, W. H. and Kamanā, K. 2001. ‘ “Mai Loko Mai O Ka ‘I’ini”: proceeding
from a dream: the ‘Aha Pūunana Leo connection in Hawaiian language
revitalization’, in L. Hinton and K. Hale (eds), pp. 147–176.
Wilson, W. H. and Kamanā, K. 2011. ‘Insights from Indigenous immersion
in Hawai’i’, in D. J. Tedick, D. Christian and T. Williams Fortune (eds),
Immersion Education: Practices, Policies, Possibilities, pp. 36–57. Bristol, UK:
Wilson, W. H. and Kamanā, K. 2013. ‘E Paepae Hou ‘Ia Ka Pōhaku: reset the
stones of the Hawaiian house platform’, in L. Hinton (ed.), pp. 101–117.
Wong, L. 1999. ‘Authenticity and the revitalization of Hawaiian’,
Anthropology and Education Quarterly 30, 94–115.
Yamada, R.-M. 2007. ‘Collaborative linguistic fieldwork: practical applica-
tion of the empowerment model’, Language Documentation and Conservation
1, 257–282.
Yamada, R.-M. 2011. ‘Integrating documentation and formal teaching of
Kari’nja: documentary materials as pedagogical materials’, Language
Documentation and Conservation 5, 1–30.
Yamamoto, A., Brenzinger, M. and Villalón, M. E. 2008. ‘A place for all
languages: on language vitality and revitalization’, Museum International
60, 60–70.

16
first language attrition
Bregtje Seton and Monika S. Schmid
16.1 Introduction
The multi-competence approach views bilingual development as a wholis-

tic process that impacts not only on the linguistic system which is being
acquired but on other languages that are already established in the mind/
brain (Cook 2012). This perspective implies that the process commonly
referred to as first language attrition – the changes to linguistic skills or
language proficiency under conditions of reduced use – should be seen as
an essential component of this wider picture. The assumption that bilin-
gual development ‘involves the whole mind of the speaker, not simply
their first language (L1) or their second’ (Cook 2012, p. 3768) puts devel-
opments and changes which occur in the first language while another is
being learned or used on an equal footing with the development of the
language that is being acquired. This status of processes of change in the
first language, however, is not reflected in present-day linguistic research,
with investigations of and insights into L1 attrition still lagging far behind
the multitude of studies of second language (L2) development.
The present contribution will give an overview of research in the field of
first language attrition in a migration setting, and try to integrate those
findings into the overall multi-competence framework. We would like to
point out that, despite the fact that the term attrition is often perceived to
imply negative connotations or collocations (cf. war of attrition), we do not
use it here with any evaluative implication. On the contrary, in the same
way that multi-competence approaches aim to consider L2 users in their
own right and deny a special status to the native speaker on the assump-
tion that ‘it is the users’ own language that matters’ (Cook 2012, p. 1), the
variety used by the attriter is not to be seen as inferior, reduced or deterio-
rated: it is simply a system which coexists in the mind/brain, and thus

Multi-competence and first language attrition 339
within a larger ‘language supersystem’ with another (possibly dominant)

language. We thus feel that, while the original label attrition may have
been somewhat unfortunately and inappropriately chosen,1 it has become
such an established term in the intervening years (with more than 5,000
hits on Google Scholar) that it would be counterproductive to change the
nomenclature at this point.
The most important contribution that investigations of attrition can
make to our understanding of multilingual development is that such
development is neither linear nor unidirectional and that it often eludes
our predictions. In other words, language users do not invariably progress
in an orderly, structured and homogeneous way from a state at which they
had no knowledge of a language through a series of stages, at each of
which their knowledge of the language is linearly greater than it was at
the previous one, converging towards (but probably never arriving at) the
idealized monolingual target variety. Proficiency, complexity, accuracy,
fluency, how comfortable a speaker is with the language, his or her
attitudes towards it, and so on, fluctuate throughout the language user’s
lifespan and in all of his or her languages, and these developments are
conditioned by many factors which interact in complex and often unpre-
dictable ways.
With the growing body of research on the brain, there are potentially
many insights to be gained from the perspective of neuroscience. An
increase in knowledge about the brain can lead to a refinement and
redefinition of the underlying models of linguistic knowledge. Thus the
goal of the present chapter is to re-evaluate the existing attrition literature
from a multi-competence perspective and to arrive at a more integrated
picture of language attrition on the one hand and language acquisition on
the other. We will conclude the survey with a brief investigation of the
importance of neurophysiological models of language processing.
16.2 Cross-linguistic interaction and the question

of the baseline
Cook (2009) distinguishes different groups of L2 users, categorized accord-

ing to the function which the L2 plays for them – for example, migrant
communities using the L2 in their environment, speakers for whom the
L2 has a limited function such as a ritualistic/religious one or a profes-
sional one, heritage speakers, and so on. These types of bilingualism have
an impact on the characteristics of the interlanguage and the amount of
bidirectional cross-linguistic interaction, but it is safe to assume that all
bilinguals experience such phenomena to some extent. For approaches
which focus on the way in which bilinguals use their languages as varieties
in their own right, as opposed to attempting to identify deficiencies with
respect to an idealized and unrealistic monolingual baseline, it is these

340 B R EGTJ E SETO N AN D M O N I K A S . SC H M I D
phenomena of interaction between all of a multi-competent user’s

languages that are of interest.
Such an approach is of particular importance within the context of
investigations that attempt to identify whether there are inherent or
maturational limits to attainment in second language development.
From a theoretical perspective which acknowledges the interaction of all
linguistic systems known by the user, attempts to assess the performance
of L2 speakers against that of monolinguals in order to identify any
residual ‘deficits’ – such as the studies by Abrahamsson and Hyltenstam
(2009), Bongaerts et al. (1997) or Hopp (2007), to name but a few – are
inherently meaningless in that they may be comparing apples and
oranges. It has long been known that bilingual users experience cross-
linguistic traffic in all of their linguistic systems. Initially (at low levels of
experience with the L2), there will be more accommodation of the L2 to the
L1, but with increased experience, this status may apparently shift, and
traffic in the other direction can increase. This was, for example, found in
an early study by Flege (1987), who investigated phonetic production of
French–American English and American English–French bilinguals. His
investigation of Voice Onset Time (VOT) among these speakers revealed
that all bilinguals differed from the monolingual controls in both their
languages, in that their VOTs had shifted towards the values set by the
other language (although the values for both languages were still clearly
discriminated, i.e., the two phonetic systems had not merged or con-
verged). In the L1, this shift was smallest for the users who had had only
nine months’ exposure to the L2. On the other hand, for that group of
low-experience L2 users, the VOT values of the L2 were furthest away from
native production. In long-term, immersed bilinguals, cross-linguistic
adaptation in the L1 increased but became reduced in the L2.
Since this and other studies clearly indicate that the native language of
a multilingual language user undergoes measurable change after only a
few months’ exposure to the L2 (and other studies, such as Chang 2012,
have found similar effects after as little as six weeks), it seems absurd to
expect any L2 user to arrive at a state where s/he will become indistinguish-
able from a monolingual native, or to derive theoretical explanations on
the nature of second language learning from their ‘failure’ to do so. As
Hopp and Schmid (2013) and Schmid (2014) have recently argued, more
insight into the way in which language is acquired, processed, and used
can be gained by comparing bilinguals who use the same languages but
were exposed to them in a different sequence (e.g. English–German and
German–English bilinguals). Such investigations have the potential of
revealing whether the mechanisms underlying language learning do
indeed change during the lifespan, what the impact is of factors such
as entrenchment (reinforcement of the memory trace through repetition,
i.e. a frequency effect) and competition, and how different language
systems at varying levels of development interact in the human mind.

16.3 Predictors of first language attrition
The most striking characteristic that sets bilinguals apart from monolin-
guals is a larger amount of variability in performance. Investigations of
language attrition have often struggled to find adequate descriptive or
explanatory frameworks to account for the fact that one attriter may
exhibit more cross-linguistic influence than another. In early investiga-
tions it was often assumed, for example, that the process of language loss,
change or reconstruction would exhibit a pattern that would be the
reverse of the acquisitional process (as e.g. predicted by Jakobson 1941,
in what has come to be known as the ‘Regression Hypothesis’) or that there
would be a straightforward relationship between the amount of exposure
and the degree of deterioration (the ‘use it or lose it’ tenet, e.g. Andersen
1982). However, more recent empirical investigations have shown that
the picture is more complex and that many factors not only play a role in
the process but may also interact. For example, Keijzer (2007, 2010) finds
interesting but limited connections between order of acquisition and
attrition of some morphological features, but also demonstrates changes
that can be ascribed to cross-linguistic influence which is absent in first
language acquisition. Where the – apparently obvious and simplistic –
prediction that lack of input and use will lead to attrition is concerned,
various investigations find no clear or straightforward relation between
the two (de Bot, Gommans and Rossing 1991; Köpke 1999, 2007; Schmid
2007; Schmid and Dusseldorp 2010; Schmid 2011).2
Similarly, it has been shown that attrition over long periods of time is
not a linear process, but that the bulk of changes to L1 accuracy and
accessibility appears to take place relatively early on, probably within
the first decade of emigration. The strongest evidence in this respect
comes from an investigation by de Bot and Clyne (1989, 1994), who
observed no (further) change in long-term emigrants at two testing
moments even though these were separated by a fifteen-year gap. This
suggests that the key factor in the attritional process may not only be
scarcity of input or use, but also that the effort devoted to acquiring
another language (which is presumably highest in the first years after
arrival in the new country) may play a more important role, along the
lines of the argument made by Herdina and Jessner (2002) that the effort to
acquire a language (language acquisition effort) has an influence on the
effort to maintain another already acquired language (language mainte-
nance effort). In a similar vein, more personal factors also appear to play a
role. For example, Schmid (2002, 2004) shows that emotional factors such
as traumatic experiences, attitudes and identification can considerably
influence the attrition process.
The only clear-cut and unambiguously important factor in determining
language attrition appears to be age of arrival. Those few studies which

investigate both pre- and post-puberty migrants (Ammerlaan 1996, 1997;

Pelc 2001; Bylund 2009) find age at onset to be the most important
predictor, while in populations which migrated above the age of 12 the
significance of this factor disappears entirely (Köpke 1999; Schmid 2002).
The most extreme instances of L1 attrition can be found among interna-
tional adoptees whose L1 seems to have literally been ‘erased’ from the
brain (Pallier et al. 2003; Ventureyra, Pallier and Yoo 2004). In this respect,
the concept of ‘incomplete acquisition’ is often invoked (e.g. Montrul
2008). However, as was recently pointed out (Bylund 2009; Schmid 2013),
the age range of 10 to 12 appears to be an important threshold for suscept-
ibility to reconstruction, change or deterioration for a variety of linguistic
skills which are mastered at quite different stages of the acquisition
process, such as nativeness in pronunciation (Yeni-Komshian, Flege and
Liu 2000), case-marking (Schmitt 2010), tense, mood and aspect (Silva-
Corvalán 1994), lexical naming (Ammerlaan 1996) or global proficiency
(Hakuta and D’Andrea 1992).
Assessing the impact of the factors listed above which act as predictors
in the language attrition process is furthermore complicated by the fact
that many of them may interact in complex and non-linear ways. For
example, the factors of Length of Residence (LoR) and Age of Emigration
(AoE) are often correlated (as e.g. in Ammerlaan 1996), since the longer the
participants have been living in the host country, the younger they usually
were when they emigrated. Hence, length of residence cannot always be
interpreted independently of age of emigration. Nevertheless, there are
studies that do find a significant effect of length of residence with late
attriters, i.e. emigrants after age 17 (Jaspaert and Kroon 1989). Other
studies only find a negative effect of length of residence when there is
little use of the language (de Bot et al. 1991; Soesman 1997), suggesting an
interaction of these factors.
The complex, non-linear interplay of predictors on the process of
language attrition has recently been approached from the theoretical
framework of Dynamic Systems Theory (de Bot, Chapter 6, this volume;
Opitz, de Leeuw and Lubińska 2013) which views language development as
a dynamic system in which multiple factors influence one another differ-
ently in different stages of development. It is clear that more research on
large populations and sophisticated statistical modelling is needed in
order to gain further insight into the role and interaction of predictor
factors for attrition.
As this short review shows, many of the factors which play a role in
bilingual development in general have been addressed and investigated
in the process of L1 attrition. The starting hypothesis for these investiga-
tions has usually been the assumption that what matters for L2 acquisi-
tion will play a similar role in L1 attrition. The actual data, however,
have usually failed to validate such straightforward hypotheses and the
question of what causes one individual language user to experience more

cross-linguistic influence in their L1 than another remains unanswered.

In this context it should be pointed out that the processes of acquisition
and attrition are closely linked with each other, and that more insight is
potentially to be gained by investigations that abandon the typical
approach of studying one or the other of these developments and take a
wider approach which includes the interaction of the languages and the
processes of change (e.g. Schmid and Köpke 2007; Schmid 2014).
16.4 Neurophysiological explanations
In order to gain more insight into the processes and factors that condition
language attrition in the individual language user, it is important to take
into account the neurocognition of linguistic knowledge, language use
and processing, and bilingualism. Of the neurophysiological processes
underpinning cognition in general and language processing and use in
particular, the most straightforward concerns the role of neuronal activity
and the constant rewiring of synapses depending on the activation of the
neurons. The more activation there is between two neurons, the higher
this amplitude of the activation between the neurons becomes and the
stronger the activation will be. This goes back to the Hebbian idea: cells
that fire together wire together (Hebb 1949).
Of course, it is not simply the firing of neurons that takes place in the
brain. There is, for example, a difference between excitatory and inhibi-
tory cells that gives rise to a more complex interplay within and between
different neural networks (Green 1998; Köpke 2007). Moreover, many
different hormones and neurotransmitters are involved in these neural
processes, and they bring about neurochemical reactions that are indis-
pensable for the brain’s functions (Lynch 2004; Ullman 2008). The storage
and consolidation of memory is assumed to happen mostly through a
strengthening of synaptic connections between neurons (e.g. Squire
1985) that are activated and re-activated. Certain parts of the brain, such
as the hippocampus and the amygdala, have been found to be actively
involved in steering these memory formations. Additionally, sleep is
thought to be an important factor in this ability to consolidate information
and transfer newly acquired knowledge into long-term memory (Stickgold
2005). This is also the case for language learning, as can be deduced from
studies finding better retention of language after episodes of sleep (Fenn,
Nusbaum and Margoliash 2003). An in-depth review of all the neural
processes involved is beyond the scope of the present chapter, all the
more so since the exact processes that give rise to memory consolidation
are still unclear (Lynch 2004; Stickgold 2005). We will therefore focus here
on a neuronal activity-based account of the consolidation of the language.
In fact, many language models, such as simple recurrent network models
(Elman 2011), parallel distributed models (McClelland 2011) and other

connectionist or neural network models (Seidenberg 1997), are based on

this simplified notion of neural activation.
On this view of neuronal activation, every time a neuron produces
electrical activity, an electrochemical trace is left along the path of the
activity. If this same path is travelled again from one neuron to another,
this event will be facilitated. If, on the other hand, there is no recurrence
then the trace will fade. Through this process of entrenchment, the access
to certain items becomes easier as they are called upon more frequently.
Entrenchment has often been invoked in the context of the formation of
cognitive routines (e.g. Langacker 2000, 2002), a process whereby certain
functions become more and more automatized and entrenched through
consecutive use.
Of course, a lack of entrenchment in the first language could very well
explain why attriters with early ages of emigration show more signs of
attrition, and why AoE does not appear to play such a large role at later
ages of emigration. In a way, entrenchment is confounded with AoE and
these two factors are then difficult to tease apart. For the acquisition of an
L2 however, this idea of entrenchment does not explain why it is often
found to be easier to learn an L2 at earlier ages as opposed to later on. A
solid neurophysiological explanation for this is still lacking, but so far the
strongest indicator seems to be synaptic over-production during the initial
stages of development of the human brain. Different cortices have differ-
ent peaks in the number of synapses available, ranging from the age of 4 to
the age of 20 (Neville and Bavelier 2002). During this over-production of
synapses, the extensive growth and elimination of axons and dendrites,
and the forming and breaking of the connections between pre- and post-
synaptic neurons leads to the development and stabilization of neural
circuits (Knudsen 2004). A significant reduction of synapses plus an
increase in stabilized synapses may be contributing to the critical or
sensitive period that is often referred to in general learning, as well as –
for our purposes – in L2 acquisition research. Nevertheless, the ongoing
rearrangement of synapses after this period still allows for a certain sensi-
tivity to change and can therefore explain why it is possible for learning to
take place at later ages and why second language learners can sometimes
be very successful and native-like in their second language (Bavelier and
Neville 2002; Perani and Abutalebi 2005).
The entrenchment account is important in the context of L1 attrition: no
other abstract system of knowledge is used as frequently and regularly as
that of the native language in a monolingual. It has been suggested that,
due to the ‘over-rehearsal’ of this system, a saturation point of entrench-
ment may eventually be reached, after which frequent reactivation is no
longer necessary to prevent the memory traces from fading, and that this
may explain the lack of evidence found in support of the ‘use it or lose it’
tenet for L1 attrition (Schmid 2007). On the other hand, as the speaker
becomes bilingual, memory traces will become established for the L2

which are extremely similar to those of the L1, and as proficiency and
frequency of use of that language increase, it will become a formidable
competitor. In some cases, the stronger L2 connections will inadvertently
serve the speaker with L2 items, structures or sounds before the L1 can be
activated. In this manner competition from the L2 may slow down or
prevent access to the L1. Such a picture suggests that cross-linguistic
interference and/or problems of accessibility in L1 attrition are not so
much linked to actual ‘language loss’ (as would be the case if the memory
traces had indeed eroded) as to the difficulty in accessing them, since this
activation requires considerable inhibitory effort (Köpke 1999; Hulsen
2000). Such a scenario appears likely in the context of evidence showing
that comprehension is typically less affected by problems of lexical access
than production, as Hulsen’s (2000) study of lexical naming versus lexical
decision illustrated, that apparently ‘forgotten’ items are (re-)acquired
faster than new ones (the ‘savings paradigm’: de Bot, Martens and
Stoessel 2004) and that the spoken L1 of attriting language users is char-
acterized by an increase in disfluency markers (Schmid and Beers
Fägersten 2010). A potential explanation for the production/comprehension
asymmetry is that a correct item from the input can be accessed by firing
neurons ‘recognizing’ it (top-down processing), while finding the correct
item through the maze of neurons that are firing in connection to the
meaning of the item (bottom-up processing) is more difficult.
On this account a more important role for the process of language
attrition is attributed to competing neuronal activity in the stronger L2
than to an actual degradation of the connections which represent the L1.
In other words, cross-linguistic interference and problems of accessibility
in the attriting language are due to a lack of inhibitory mechanisms as
opposed to an insurmountable rise of the Activation Threshold (Levy et al.
2007; Linck, Kroll and Sunderman 2009). Evidence in support of this view
comes from the fact that the only predictor linked with language use
which has been shown to be consistently important in preventing attrition
phenomena is the use of the L1 for professional purposes (Schmid 2007;
Schmid and Dusseldorp 2010). In such (formal) contexts, code-switching
and code mixing is usually not considered appropriate – as opposed to the
informal interactions with family members and friends in the country of
migration which are usually assessed and quantified in attrition studies.
Schmid (2007) thus hypothesizes that speakers who regularly use their L1
in the professional domain have more practice at resisting any intrusion
from their highly active L2 and concludes that L1 attrition phenomena are
predominantly a matter of the efficiency of processes linked to inhibition
as opposed to activation problems.
Such an account may also go some way towards explaining the phenom-
enon of ‘language reversion’ that is sometimes reported by aging migrant
populations. An early account of this phenomenon is presented by Clyne
(1977) and de Bot and Clyne (1989), who find that a number of their

participants reported that, as they grew older, their L2 skills appeared to

deteriorate and the L1 was reasserting itself. This account, however, was
based solely on anecdotal evidence and has never been substantiated in
empirical research. Clyne himself refers to reversion as a ‘myth’ later on
(Clyne 2011). In this context, it is important to note that one of the most
prevalent changes in cognitive ability that has been linked with the nor-
mal aging process is an attenuation of inhibitory processes (Burke 1997;
Radvansky, Zacks and Hasher 2005; Burke and Osborne 2007; Burke and
Shafto 2008). It may be that the deterioration of inhibitory efficiency
among older attriters can lead to more extensive code-switching and
code mixing in both languages, but that, so far, they were mainly used to
experiencing this in the L1. Such an unexpected intrusion of L1 elements
or items in the L2 may then be interpreted in the context of the strong and
widespread language reversion myth: Clyne (2011) reports that virtually
all elderly migrants that he spoke to confidently expect this to happen to
them at some point in the future (refreshingly, he cites an 85-year-old who
says that she knows she will lose her L2 and revert to her L1 ‘when I grow
old’ (p. 215).
It seems clear, therefore, that investigations of L1 attrition which take
into account neurophysiological processes may not only be able to find
explanations for some of the puzzling and unexpected phenomena which
have often been encountered in previous investigations of attrition, but to
shed more light on bilingual development in general. The assumption
that effects of the second language on the first are due mainly to a lack
of inhibitory control, and not so much to a degradation of the memory
traces underpinning the L1, is, of course, very compatible with multi-
competence approaches: as languages come to coexist within the same
mind/brain, they will interact with each other, and where such interaction
is not desired, the language that the user does not wish to select has to be
inhibited. This inhibition process becomes more and more effortful the
more deeply entrenched the linguistic system in question has become, and
that is what leads to the process that is commonly (but misleadingly)
labelled as first language attrition.
16.5 Neurocognitive approaches to language attrition
As was pointed out above, what little we know to date about the neuro-
physiological processes underpinning the language attrition process is
derived from general investigations of the brain and/or language proces-
sing, not from studies focusing specifically on language attrition.
Bilingualism research to date overwhelmingly favours investigations
of the second, the later-learned or the weaker language, and very few
studies investigate changes to linguistic systems already established at
the onset of L2 learning. This is even more true for investigations of the

neurocognitive underpinnings of bilingualism: there is a large body of

neuroimaging research of L2 processing, while there is hardly any direct
evidence concerning the effects that bilingualism has on the L1.
In this context, too, converging evidence from developments which take
place in the L1 in situations of intense bilingual use and exposure may help
us gain a deeper understanding of language acquisition and processing.
For example, while it is commonly accepted that L1 and L2 processing rely
on the same neural mechanisms and activate similar regions in the brain,
the distribution of activation in L2 speakers is also affected by factors such
as proficiency levels, age at learning, and amount of exposure (Perani et al.
2003; Perani and Abutalebi 2005). For the native language, these factors are
generally assumed to be constant because monolingualism is considered
the default. Moreover, there are no studies investigating if the neural
underpinnings of (first) language processing may change when the indivi-
dual speaker becomes bilingual. In other words, there may be changes in
brain activation patterns when extensive bilingualism leads to a decrease
in overall L1 proficiency or a dramatic reduction in L1 exposure and use.
Similarly, investigations using other neuroimaging methods, such as
Event-Related Potentials (ERPs) and eye tracking, may help elucidate to
what extent bilinguals rely on different and differential features of the
language. There is evidence from studies using ERPs, for example, that as
L2 learners become more proficient, their processing strategies change to
some extent: in low-proficiency learners, one sometimes finds brain
responses that are more typical for efforts associated with semantic inte-
gration (the N400 component) when processing violations associated with
grammatical features such as gender concord (Osterhout et al. 2004).
Higher-proficiency learners, on the other hand, tend to exhibit the P6003
typically also found among native speakers when they encounter such
structures, although this response is often somewhat delayed. Such find-
ings are in line with proposals that assume that language learners, parti-
cularly when they are not yet very experienced, use lexical cues more than
native speakers do, who have fully automatized their grammatical proces-
sing system (e.g. the ‘Shallow Structure Hypothesis’: Clahsen and Felser
2006). It would be very interesting to see what happens to long-term
attriters in this type of paradigm. One possible scenario might be that, as
automatized processing slows down or fails due to extended non-use, the
speaker may fall back on declarative representations (this intriguing
possibility was suggested by Osterhout 2014 for second language attrition).
On the other hand, it is possible that automatic processing is preserved
better in language attrition than declarative representations, but that it
may become somewhat delayed or less efficient.
None of these questions have been systematically addressed by
research on the neurocognitive aspects of bilingualism. Investigations
into the neurocognition of the first language in long-term immersed
bilinguals may help us gain a deeper understanding of multilingual

language processing which can also shed light on questions of second

language learning, as it can improve our understanding of how cross-
linguistic interaction affects language processing in native speakers.
This makes it possible to control for the ‘bilingualism factor’, as
opposed to the more common comparisons of bilingual apples and
monolingual oranges.
16.6 Conclusion
Any investigation of bilingual development which takes seriously the

proposal of multi-competence, that is, a model of linguistic knowledge
where all languages coexist within a larger supersystem and affect each
other, can be considered incomplete if it merely looks at such changes in
languages that are learned later in life. For simultaneous and early bilin-
guals, it is already often the case that both languages are investigated, but
where late bilinguals are concerned, the native language has, so far, been
largely neglected. On the other hand, very few studies which do investigate
L1 attrition are also interested in what is happening with the L2. We
suggest that such integrated approaches are needed in order to move
forward our understanding of bilingual development. In addition, a per-
spective which takes into account not only linguistic behaviour but also its
neurocognitive underpinnings may shed more light on the puzzling
phenomena that we are observing thus far.
Notes
1. Despite numerous efforts on the part of the authors and other col-
leagues, we have been unsuccessful in tracking down the first usage
of the term ‘attrition’ for this kind of linguistic development. The term
‘language attrition’ seems to have been used for the phenomenon of
language change (Friedrich 1971), as well as for language instruction
dropout (Papalia 1970). It seems likely that either of these uses may
have been a precursor for the term, which has been in common use
since the 1980s (e.g. at the 1980 UPenn symposium The Loss of Language
Skills, convened by Barbara Freed and Richard Lambert; see Köpke and
Schmid 2004).
2. All of these studies acknowledge that quantity and quality of use
are among the most difficult factors to measure in any situation of
bilingual development, in particular since they inevitably seem to rely
on subjective self-ratings by the participants themselves. They do take
all possible care to arrive at sound and reliable estimates on the basis of
these self-reports, but it is, of course, possible that this input factor
produces less clear results due to errors of measurement, as opposed to

more straightforward personal background factors such as age or

length of residence.
3. The N400 is a negative-going deflection in the ERP that is generally
found to peak around 400 ms after the presentation of the stimulus.
This component is found to be larger for lexical items that are less
frequent, or semantically less expected or implausible (e.g. spread the
bread with socks as opposed to spread the bread with butter in Kutas and
Hillyard 1980). It is therefore generally associated with semantic inte-
gration, but it has recently been claimed to be a marker of lexical access
instead of semantic integration (Brouwer, Fitz and Hoeks 2012).
The P600 is a positive-going deflection around 600 ms after the onset
of the stimulus. The onset may be somewhat earlier, and it may peak
around 900 ms, since it is quite a large component. It is generally
associated with reanalysis of syntactic violations, but in the recent
account of Brouwer et al. (2012) it is claimed to be a marker of seman-
tic integration.
References

tiny’, Language Learning 59 (2), 249–306.
Ammerlaan, T. 1996. ‘You get a bit wobbly . . .’: Exploring Bilingual Lexical
Retrieval Processes in the Context of First Language Attrition. Unpublished
doctoral dissertation, Katholieke Universiteit Nijmegen.
Ammerlaan, T. 1997. ‘ “Corrosion” or “loss” of immigrant Dutch in
Australia: an experiment on first language attrition’, in J. Klatter-
Folmer and S. Kroon (eds), Dutch Overseas: Studies in Maintenance and
Loss of Dutch as an Immigrant Language, pp. 69–97. Tilburg University
Press.
Andersen, R. W. 1982. ‘Determining the linguistic attributes of language
attrition’, in R. Lambert and B. Freed (eds), The Loss of Language Skills,
pp. 83–118. Rowley, MA: Newbury House.
Bavelier, D. and Neville, H. 2002. ‘Cross-modal plasticity: where and
how?’, Nature Reviews Neuroscience 3, 443–452.
Bongaerts, T., van Summeren, C., Planken, B. and Schils, E. 1997. ‘Age and
ultimate attainment in the pronunciation of a foreign language’, Studies
in Second Language Acquisition 19, 447–465.
Brouwer, H., Fitz, H. and Hoeks, J. 2012. ‘Getting real about semantic
illusions: rethinking the functional role of the P600 in language com-
prehension’, Brain Research 1446, 127–143.
Burke, D. 1997. ‘Language, aging, and inhibitory deficits: evaluation of a
theory’, Journals of Gerontology, Series B: Psychological Sciences and Social
Sciences 52 (6), 254–264.

Burke, D. and Osborne, G. 2007. ‘Aging and inhibition deficits: where are
the effects?’, in D. S. Gorfein and C. M. MacLeod (eds), On the Place of
Inhibitory Processes in Cognition, pp. 63–183. Washington, DC: American
Psychological Association Press.
Burke, D. and Shafto, M. A. 2008. ‘Language and aging’, in F. I. M. Craik and
T. A. Salthouse (eds), The Handbook of Aging and Cognition, pp. 373–443.
New York: Psychology Press.
Bylund, E. 2009. ‘Maturational constraints and first language attrition’,
Language Learning 59 (3), 687–715.
249–268.
Clahsen, H. and Felser, C. 2006. ‘Grammatical processing in language
learners’, Applied Psycholinguistics 27, 3–42.
Clyne, M. 1977. ‘Bilingualism in the elderly’, Talanya 4, 45–56.
Clyne, M. 2011. ‘Bilingualism, code-switching and aging: a myth of attri-
tion and a tale of collaboration’, in M. S. Schmid and W. M. Lowie (eds),
Modeling Bilingualism: From Structure to Chaos, pp. 201–220. Amsterdam and
Cook, V. J. 2009. ‘Language user groups and language teaching’, in
Cook, V. J. 2012. ‘Multi-competence’, in C. A. Chapelle (ed.), The
Encyclopedia of Applied Linguistics, pp. 3768–3774. Oxford: Blackwell.
De Bot, K. and Clyne, M. 1989. ‘Language reversion revisited’, Studies in
Second Language Acquisition 11, 167–177.
De Bot, K. and Clyne, M. 1994. ‘A 16 year longitudinal study of language
attrition in Dutch immigrants in Australia’, Journal of Multilingual and
Multicultural Development 15 (1), 17–28.
De Bot, K., Gommans, P. and Rossing, C. 1991. ‘L1 loss in an L2 environ-
ment: Dutch immigrants in France’, in H. W. Seliger and R. M. Vago (eds),
First Language Attrition, pp. 87–98. Cambridge University Press.
De Bot, K., Martens, V. and Stoessel, S. 2004. ‘Finding residual lexical
knowledge: the “savings” approach to testing vocabulary’, International
Journal of Bilingualism 8 (3), 373–382.
Elman, J. L. 2011. ‘Lexical knowledge without a lexicon?’, The Mental Lexicon
6 (1), 1–33.
Fenn, K. M., Nusbaum, H. C. and Margoliash, D. 2003. ‘Consolidation
during sleep of perceptual learning of spoken language’, Nature 425,
614–616.
Friedrich, P. 1971. ‘Dialectal variation in Tarascan phonology’, International
Journal of American Linguistics 37 (3), 164–187.

Green, D. W. 1998. ‘Schemas, tags and inhibition’, Bilingualism: Language

Hakuta, K. and D’Andrea, D. 1992. ‘Some properties of bilingual mainte-
nance and loss in Mexican background high-school students’, Applied
Linguistics 13 (1), 72–99.
Hebb, D. O. 1949. The Organization of Behavior: A Neurophysiological Approach.
New York: Wiley.
Hopp, H. 2007. Ultimate Attainment at the Interfaces in Second Language
Acquisition: Grammar and Processing. Doctoral dissertation, University of
Groningen.
Hopp, H. and Schmid, M. S. 2013. ‘Perceived foreign accent in L1 attrition
and L2 acquisition: the impact of age of acquisition and bilingualism’,
Applied Psycholinguistics 34 (2), 361–394.
Hulsen, M. 2000. Language Loss and Language Processing: Three Generations of
Dutch Migrants in New Zealand. Unpublished doctoral dissertation,
Katholieke Universiteit Nijmegen.
Jakobson, R. 1941. Kindersprache, Aphasie und allgemeine Lautgesetze. Uppsala:
Almqvist & Wiksell.
Jaspaert, K. and Kroon, S. 1989. ‘Social determinants of language loss’,
Review of Applied Linguistics (I.T.L.) 83/84, 75–98.
Keijzer, M. 2007. First Language Attrition: An Investigation of Jakobson’s
Regression Hypothesis. Doctoral dissertation, Vrije Universiteit
Amsterdam.
Keijzer, M. 2010. ‘The regression hypothesis as a framework for first
language attrition’, Bilingualism: Language and Cognition 13 (1), 9–18.
Knudsen, E. 2004. ‘Sensitive periods in the development of the brain and
behavior’, Journal of Cognitive Neuroscience 16 (8), 1412–1425.
Köpke, B. 1999. L’Attrition de la Première Langue Chez le Bilingue Tardif:
Implications pour l’Étude Psycholinguistique du Bilinguisme. Unpublished
doctoral dissertation, Université de Toulouse-Le Mirail.
Köpke, B. 2007. ‘Language attrition at the crossroads of brain, mind, and
society’, in B. Köpke, M. S. Schmid, M. Keijzer and S. Dostert (eds),
pp. 9–37.
Köpke, B. and Schmid, M. S. 2004. ‘Language attrition: the next phase’, in
M. S. Schmid, B. Köpke, M. Keijzer and L. Weilemar (eds), First Language
Attrition: Interdisciplinary Perspectives on Methodological Issues, pp. 1–43.
Köpke, B., Schmid, M. S., Keijzer, M. and Dostert, S. (eds) 2007. Language
Attrition: Theoretical Perspectives. Amsterdam: John Benjamins
Kutas, M. and Hillyard, S. A. 1980. ‘Reading senseless sentences: brain
potentials reflect semantic incongruity’, Science 207 (4427), 203–205.
Langacker, R. W. 2000. Grammar and Conceptualization. Berlin: Mouton De
Gruyter.

Langacker, R. W. 2002. Concept, Image, and Symbol: The Cognitive Basis of

Grammar, second edition. Berlin: Mouton De Gruyter.
Levy, B. J., McVeigh, N. D., Marful, A. and Anderson, M. C. 2007. ‘Inhibiting
your native language: the role of retrieval-induced forgetting during
second-language acquisition’, Psychological Science 18 (1), 29–34.
Linck, J. A., Kroll, J. F. and Sunderman, G. 2009. ‘Losing access to the native
language while immersed in a second language: evidence for the role
of inhibition in second-language learning’, Psychological Science 20 (12),
1507–1515.
Lynch, M. A. 2004. ‘Long-term potentiation and memory’, Physiological
Review 84, 87–136.
McClelland, J. L. 2011. ‘Memory as a constructive process: the parallel
distributed processing approach’, in S. Nalbantian, P. Matthews and
J. L. McClelland (eds), The Memory Process: Neuroscientific and Humanistic
Perspectives, pp. 129–151. Cambridge, MA: MIT Press.
Montrul, S. 2008. Incomplete Acquisition in Bilingualism: Re-examining the Age
Factor. Amsterdam and Philadelphia: John Benjamins.
Neville, H. and Bavelier, D. 2002. ‘Human brain plasticity: evidence
from sensory deprivation and altered language experience’, in
M. A. Hofman, G. J. Boer, A. J. G. D. Holtmaat, E. J. W. van Someren,
J. Verhaagen and D. F. Swaab (eds), Plasticity in the Adult Brain: From
Genes to Neurotherapy, pp. 177–188, vol. 138 of Progress in Brain Research.
Amsterdam: Elsevier.
Opitz, C., de Leeuw, E. and Lubińska, D. (eds) 2013. ‘Dynamics of first
language attrition across the lifespan’, International Journal of
Bilingualism 17 (6), 667–674.
Osterhout, L. 2014. Event-related brain potentials as metrics of non-native
language learning and loss. Paper presented at the symposium Second
Language in the Brain, University of Greenwich, 4 October 2014.
Osterhout, L., McLaughlin, J., Kim, A., Greenwald, R. and Inoue, K. 2004.
‘Sentences in the brain: event-related potentials as real-time reflections
of sentence comprehension and language learning’, in M. Carreiras and
C. J. Clifton (eds), The On-line Study of Sentence Comprehension: Eyetracking,
ERP, and Beyond, pp. 271–308. New York: Psychology Press.
Pallier, C., Dehaene, S., Poline, J. B., LeBihan, D., Mehler, J. and Dupoux,
E. 2003. ‘Brain imaging of language plasticity in adopted adults: can a
second language replace the first?’, Cerebral Cortex 13, 155–161.
Papalia, A. 1970. ‘A study of attrition in foreign language enrolments in
four suburban public schools’, Foreign Language Annals 4 (1), 62–67.
Pelc, L. 2001. L1 Lexical, Morphological and Morphosyntactic Attrition in Greek–
English Bilinguals. Doctoral dissertation, CUNY, New York.
Perani, D. and Abutalebi, J. 2005. ‘The neural basis of first and second
language processing’, Current Opinion in Neurobiology 15 (2), 202–206.
Perani, D., Abutalebi, J., Paulesu, E., Brambati, S., Scifo, P., Cappa, S. F. and
Fazio, F. 2003. ‘The role of age of acquisition and language usage in early,

high-proficient bilinguals: an fMRI study during verbal fluency’, Human

Brain Mapping 19 (3), 170–182.
Radvansky, G. A., Zacks, R. T. and Hasher, L. 2005. ‘Age and inhibition: the
retrieval of situation models’, The Journals of Gerontology, Series B 60 (5),
276–278.
Schmid, M. S. 2002. First Language Attrition, Use and Maintenance: The Case of
German Jews in Anglophone Countries. Amsterdam: John Benjamins.
Schmid, M. S. 2004. ‘First language attrition: the methodology revised’,
International Journal of Bilingualism 8 (3), 239–255.
Schmid, M. S. 2007. ‘The role of L1 use for L1 attrition’, in B. Köpke,
M. S. Schmid, M. Keijzer and S. Dostert (eds), pp. 135–153.
Schmid, M. S. 2013. ‘First language attrition’, Linguistic Approaches to
Schmid, M. S. 2014. ‘The debate on maturational constraints in bilingual
development: a perspective from first language attrition’, Language
Acquisition 21 (4), 386–410.
Schmid, M. S. and Beers Fägersten, K. 2010. ‘Disfluency markers in L1
attrition’, Language Learning 60 (4), 753–791.
Schmid, M. S. and Dusseldorp, E. 2010. ‘Quantitative analyses in a
multivariate study of language attrition: the impact of extralinguistic
factors’, Second Language Research 26 (1), 125–160.
Schmid, M. S. and Köpke, B. 2007. ‘Bilingualism and attrition’, in B. Köpke,
M. S. Schmid, M. Keijzer and S. Dostert (eds), Language Attrition: Theoretical
Perspectives, pp. 1–8. Amsterdam: John Benjamins.
Schmitt, E. 2010. ‘When boundaries are crossed: evaluating language
attrition data from two perspectives’, Bilingualism: Language and
Cognition 13 (1), 63–72.
Seidenberg, M. S. 1997. ‘Language acquisition and use: learning and
applying probabilistic constraints’, Science 275, 1599–1603.
Silva-Corvalán, C. 1994. Language Contact and Change: Spanish in Los Angeles.
Oxford: Clarendon Press.
Soesman, A. 1997. ‘An experimental study on native language attrition
in Dutch adult immigrants in Israel’, in J. Klatter-Folmer and S. Kroon
(eds), Dutch Overseas: Studies in Maintenance and Loss of Dutch as an Immigrant
Language, pp. 181–194. Tilburg University Press.
Squire, L. R. 1985. ‘Memory: neural organization and behavior’, in
J. M. Brookhart and V. B. Mountcastle (eds), Handbook of Physiology: The
Nervous System V, pp. 295–371. American Physiological Society.
Stickgold, R. 2005. ‘Sleep-dependent memory consolidation’, Nature 437
(7063), 1272–1278.
Ullman, M. T. 2008. ‘The role of memory systems in disorders of language’,
in B. Stemmer and H. A. Whitaker (eds), Handbook of the Neuroscience of
Language, pp. 189–198. Oxford, UK: Elsevier.

Ventureyra, V. A., Pallier, C. and Yoo, H.-Y. 2004. ‘The loss of first language
phonetic perception in adopted Koreans’, Journal of Neurolinguistics 17 (1),
79–91.
Yeni-Komshian, G. H., Flege, J. E. and Liu, S. 2000. ‘Pronunciation profi-
ciency in the first and second languages of Korean–English bilinguals’,
Bilingualism: Language and Cognition 3 (2), 131–149.

17
Cognitive consequences
of multi-competence
Panos Athanasopoulos
17.1 Introduction
Close to two centuries ago, the German philosopher Wilhelm von

Humboldt observed that ‘The learning of a foreign language . . .
[means] the gaining of a new standpoint toward one’s world-view’
(1836/1963, p. 294). In today’s global society, bilingualism (broadly
defined here as the use of more than one language in everyday life,
including second, third or multiple languages) has become a far more
prominent aspect of everyday life than ever before. Understanding
people’s multilingual capacities, and integrating linguistically diverse
perspectives may promote inter-cultural understanding and help
bridge socioeconomic, ethnic, and cultural divisions (The Radein
Initiative 2011). The framework of multi-competence provides a
ready platform to begin to trace the cognitive characteristics of the
bilingual person, and particularly to identify the specific cognitive
processes and mechanisms that mould the multi-competent mind
into a unique thinking entity. This is mainly down to two reasons.
First, the multi-competence framework emphasizes the need to study
bilinguals in their own right rather than as imperfect versions of a
monolingual native speaker ideal. For example, Cook (1991, 1992,
1999, 2002, 2003) views the person who speaks more than one lan-
guage as an independent speaker/hearer/thinker, with linguistic and
cognitive representations and abilities which are qualitatively distinct
from those of a monolingual person. This resonates with other bilin-
gualism scholars like Grosjean (1982, 1989, 1992, 1998), who has
repeatedly argued that the bilingual person is not two monolinguals
in the same body, but a unique language user with a complete lan-
guage system. These approaches shift the terms of engagement from
focusing on what L2 users cannot do and why, relative to native

356 PA N O S AT H A N A S O P O U LO S
speakers of one or the other language, to what bilinguals do differ-

ently and why, compared to those who only know one language.
Second, the multi-competence framework provides a coherent, uni-
fied account of bilingual cognition, because it assumes that language
and cognition are highly interactive (see Chapter 1, this volume). This
interaction has consequences for the whole mind of the L2 user, not
just the language component. In this way, multi-competence brings
together two related strands of bilingual cognition research, namely
bilingual conceptual representation, that is, the way concepts from the
specific languages of the bilingual person are integrated, and bilingual
executive functioning related to language-independent cognitive pro-
cessing specifically pertaining to inhibitory control, that is, the ability
to efficiently inhibit task-irrelevant distracting information in order to
attend to a specific task.
More specifically, research on bilingual cognition has proceeded along
two main lines of enquiry (Green 1998a). One line of research is con-
cerned with cognitive-general effects, and investigates cognitive abilities
related to executive control (blocking out competing information that is
irrelevant to the task at hand and focusing only on task-relevant informa-
tion). The necessity of bilinguals to focus on the relevant language while
ignoring interference from the other language (Green 1998b) enhances
these general cognitive skills in the individual, and may even offset
symptoms of dementia and other neurodegenerative diseases (Bialystok
2009). Another line of research investigates language-specific effects
on how bilinguals mentally represent reality and the world, and is
anchored on Whorf’s (1956) principle of linguistic relativity (the idea
that speakers of different languages think differently). Using two
languages with contrasting ways of cutting up the world leads bilinguals
to conceptualize the world in a unique way, dissimilar to their mono-
lingual counterparts, in a variety of domains like colour, number, and
gender (Athanasopoulos 2011).
The current paper aims to provide a synopsis of empirical investiga-
tions of both cognitive-general and language-specific effects of bilin-
gualism on cognition. I will begin by highlighting the two main
strands of research on bilingual cognition, briefly delineating their
origins and their trajectory through the history of science (Section
17.2). I will then review the experimental protocols and main findings
that each strand has yielded (Sections 17.3 and 17.4). The aim here is
not to be exhaustive, but rather to illustrate the main methods and
findings from each strand of research, and more importantly, to
synthesize findings from each strand, and to consider the extent to
which they help us characterize key processes of the interaction
between bilingual cognition and a range of variables that characterize
the multi-competent person (Sections 17.5 and 17.6).

Cognitive consequences of multi-competence 357
17.2 Approaches to studying cognition in the multi-

competent mind
Two inter-related key questions at the centre of bilingual cognition

research are:
– Do bilinguals think differently from monolinguals?

– Does learning additional languages change the way humans think?
These questions have traditionally been studied in two main schools of

thought. One approach focuses on cognitive abilities of multilinguals in
the context of general cognitive processes such as executive control. The
other approach aims to unravel the mental representations of multilin-
guals in the context of linguistic relativity research. Both schools of
thought are in a sense conceptually related, in that a basic tenet of both
is that the linguistic system is highly interactive with other areas of
cognitive processing. Both schools of thought exhibit a remarkably similar
trajectory in the history of science, spanning at least a century.
From the early nineteenth century to the 1960s, bilingualism was
regarded as detrimental to cognitive abilities. For example, Saer (1923)
compared IQ test scores of Welsh–English bilingual and monolingual
English speakers from rural Wales and concluded that bilinguals were
disadvantaged mentally compared to monolinguals: ‘the difference in
mental ability as revealed by intelligence tests is of a permanent nature
since it persists in students throughout their University career’ (Saer et al.
1924, p. 53). At around the same time, arguably the first ever investigation
of the relationship between bilingualism and thought was taking place in
Switzerland. Izhac Epstein wrote his doctoral dissertation (published in
1915, cited in Pavlenko 2011) on the mental representations of multilin-
guals, as revealed by inner speech, mental translation, and calculations.
Epstein argued that multilingualism slows down the thought process
through activation of alternative options available in other languages,
and concluded that ‘La polyglossie est une plaie sociale’ [Multilingualism
is a social ill] (p. 210, cited in Pavlenko 2011). In 1962, a study by Peal and
Lambert on the cognitive abilities of bilingual children in Canada marked a
major turning point in the history of bilingual cognition. The researchers
successfully overcame the methodological flaws of Saer’s and others’ tests
(see Cummins 1976), and further demonstrated that contrary to earlier
beliefs, bilingualism conferred an advantage in intelligence over mono-
lingualism. At around the same time, Susan Ervin reconceptualized the
way the relationship between languages and thought is operationalized
in bilingual speakers by carrying out a series of empirical studies that
investigated classification and interpretation of verbal and non-verbal
stimuli. She tested bilinguals twice on the same set of stimuli and showed
that bilinguals’ responses shifted depending on the language of testing

(Ervin-Tripp 1964). She also showed a semantic shift of colour category

prototypes towards the L2 in Navajo–English bilinguals (Ervin 1961).
While Peal and Lambert’s (1962) study was subsequently criticized on
methodological and sampling grounds, it nonetheless marked the begin-
ning of the era of additive effects of multilingualism on cognition. Thus
later research found advantages of multilingualism for creative thinking
(Laurén 1991), for metalinguistic awareness (greater awareness of linguis-
tic meaning and structure: Ianco-Worrall 1972), and more recently for
executive control (Bialystok 2009). Meanwhile, Ervin’s studies attracted
much interest initially, but empirical investigations on this line of enquiry
dwindled in the 1970s, 1980s, and 1990s mainly because the field of
linguistics was dominated at that time by the framework of Noam
Chomsky, which emphasized the universal basis of human cognition,
and assumed that language is an independent, unrelated module from
other areas of cognition.
On the eve of the new millennium, ground-breaking research chal-
lenged the decades-long dominance of the universal basis of human cogni-
tion, and provided new evidence in support of the linguistic relativity
principle (Lucy 1992; Levinson 1996; Roberson et al. 2000). These scholars
integrated methods used in experimental psychology with insights from
anthropological and cognitive linguistics, and placed the experimental
approach at the heart of the language and thought debate, emphasizing
the need to measure both verbal and, crucially, non-verbal performance in
a range of linguistic and cognitive tasks. However, there was concern
regarding the scarcity of data from multilinguals in these new approaches
to linguistic relativity, and the need to study the multilingual mind more
closely was raised (Green 1998a; Pavlenko 1999; Cook 2002). In 2006, two
studies began to redress the balance by conducting the first systematic
investigations of the relationship between grammatical structure and
thought in the bilingual person (a handful of earlier studies did consider
bilingualism, but only as an afterthought: see e.g. Boroditsky 2001). Cook
et al. (2006) and Athanasopoulos (2006) studied the effects of grammatical
number marking on categorization and perception of objects and sub-
stances in Japanese–English bilinguals. The studies revealed that bilin-
guals changed the way they categorized objects, shifting from a
primarily substance/material based categorization (typical for Japanese
monolinguals) towards a more shape-based categorization (more typical
of English monolinguals), as a function of increases in level of L2 profi-
ciency and length of immersion in the L2-speaking country. These and
subsequent studies (see Cook and Bassetti 2011, Jarvis 2011 and Pavlenko
2011 for recent collections of articles) provided key evidence for the
hypothesis that the bilingual person is not two monolinguals in the
same body, but an independent, multi-competent speaker/hearer/thinker
with a unique mental representation of the world (Grosjean 1998; Cook
2002).

17.3 Cognitive-general effects of bilingualism

on cognition
General effects on cognitive processes are thought to derive from how

bilinguals use their language systems in order to communicate. As men-
tioned earlier, studies show that bilinguals outperform monolinguals at
blocking out information that is irrelevant to the task at hand (Bialystok
2009) because they need to constantly manage competing inputs and
control production in one language (Abutalebi and Green 2008) in the
presence of automatic cross-activation between languages (Thierry and
Wu 2007).
One study by Bialystok et al. (2004) compared groups of younger and
older bilinguals and monolinguals using the Simon task. In this task
coloured stimuli are presented on either the left or the right side of a
computer screen. Each of two colours (or two pairs of colours) is associated
with a response key on the two sides of the keyboard underneath the
stimuli. Participants are instructed to press the key on the correct side.
For example, if a participant is instructed to press the red key on the left
side of the computer keyboard when they see a red square on the computer
screen, and the blue key on the right side of the keyboard when they see a
blue square on the screen, a correct ‘congruent’ response occurs when the
person presses the left key when red is presented on the left side of the
screen (and vice versa for blue). A correct ‘incongruent’ response is when
the subject presses the left key when red is presented on the right side of
the screen (and vice versa for blue). Incongruent trials have longer reaction
times, called the Simon effect. Bialystok et al. (2004) found that bilinguals
outperformed monolinguals on the Simon task: bilinguals performed the
task quicker than ‘matched’ monolinguals, and showed less interference
in the ‘incongruent’ trials (indexed by faster reaction times than mono-
linguals). Bialystok et al. (2004) focused on the variable of aging, as longer
reaction times tend to occur with aging in the Simon task. Bialystok et al.
(2004) found that the cognitive difference between monolinguals and
bilinguals was minimal in young adults, presumably because cognitive
ability is at its peak in that age group. The most interesting finding was
that bilingualism significantly reduced the age-related lower performance:
older bilinguals performed significantly better than older monolinguals.
In a further attempt to explore the complex interaction between the
cognitive effects of aging and bilingualism, Bialystok et al. (2008) tested an
older (age around 67 years) and younger (age around 20 years) group of
monolinguals and bilinguals in tasks revealing the functioning of working
memory, lexical retrieval, and conflict resolution. One such task is the
Stroop interference task. If participants have to name the colour of the
word ‘green’ printed in red ink, conflict between word meaning and ink
colour occurs, which slows down reaction time. The task employs two

types of trial. In congruent trials, the colour of the ink is the same as the
colour word (e.g. the word green written in green ink), but in incongruent
trials the colour of the ink conflicts with the colour word (e.g. the word
green written in red ink). Typically participants are slower to respond in the
incongruent condition than in the congruent condition. This is called the
Stroop effect. Bialystok et al. (2008) found that the Stroop effect was larger
for monolinguals and the higher age group. Hence a three-way interaction
between age, language, and Stroop effect was established. Younger and
older bilinguals suffered from a smaller Stroop effect. When it came to the
facilitation effect of naming a word printed in a congruent colour only the
older bilinguals exceeded their monolingual peers in the ability of making
use of that cue.
Emmorey et al. (2008) found bilingual advantages in another measure of
executive control, namely the flanker task. In this task participants are
asked to respond to a directed item (e.g. a chevron pointing to the left, <)
surrounded by distracters, which can be of the same shape as the target but
differ in one dimension such as colour (e.g. a red chevron target sur-
rounded by grey chevron distracters). In congruent trials, the distracters
point to the same direction as the target (e.g. all chevrons pointing to the
left, or all chevrons pointing to the right). In incongruent trials, the
distracter chevrons point to a different direction from the target (e.g. a
target chevron pointing left, surrounded by chevrons pointing right).
Participants are asked to indicate the direction the target chevron is
pointing to as accurately and quickly as possible. Successful performance
in the incongruent version of the task requires participants to inhibit the
directionality of the distracters in order to correctly respond to the direc-
tionality of the target. Typically performance is slower and less accurate in
the incongruent than in the congruent trials. Emmorey et al. (2008) imple-
mented this task in monolinguals, unimodal bilinguals (people who use
two spoken languages), and bimodal bilinguals (people who use one spo-
ken and one sign language). Results from these three groups elucidated the
source of executive control enhancement in bilinguals. Specifically, uni-
modal bilinguals performed faster than monolinguals and bimodal bilin-
guals. The latter group did not differ from monolinguals. This finding
traces the bilingual advantage to the unimodal bilinguals’ experience in
controlling two languages in the same modality, in contrast to bimodal
bilinguals, whose languages involve distinct modalities (speech and sign)
that can be used simultaneously without the need to inhibit one while
using the other.
The conclusion drawn by studies of this kind is that bilingual advantages
in cognitive control derive from constant practice at language selection
through inhibition of competing linguistic items during language use.
Lifelong experience in efficiently inhibiting one language while using
the other tempers the age-related decline in the efficiency of inhibitory
processing more generally. Consequently, lifelong bilingualism may

provide a partial defence against the normal decline in cognitive control

associated with aging. Indeed, Bialystok, Craik and Freedman (2007) com-
pared the age of onset of dementia in a sample of 184 individuals, half of
whom were bilingual and half monolingual. The study revealed ‘that
bilinguals showed signs of dementia four years later than the monolin-
guals – with a mean age of 71.4 and 75.5 for monolinguals and bilinguals
respectively’ (Bialystok 2009, pp. 8–9).
17.4 Language-specific effects of bilingualism on cognition
Language-specific cognitive effects arise as a result of how specific lan-

guages encode and carve up the physical world. If different languages
divide up the world in different ways, then the question arises as to how
the world is represented in the mind of bilinguals with languages that have
contrasting lexical and grammatical categories. These kinds of investiga-
tions are mainly anchored on Whorf’s principle of linguistic relativity
(Whorf 1956) that states that, since different languages have different
concepts, speakers of different languages think and reason about the
physical world differently: ‘users of markedly different grammars are
pointed by their grammars toward different types of observations and
different evaluations of externally similar acts of observation’ (Whorf
1956/1940, p. 221). The past fifteen years have witnessed a surge of empiri-
cal research on linguistic relativity. An increasing number of studies
demonstrate effects of specific lexical and grammatical categories on
specific cognitive processes such as remembering, thinking, and reason-
ing about the perceptual relationships of physical entities like colours,
objects, and events; for reviews, see Gentner and Goldin-Meadow (2003),
Roberson (2005), Casasanto (2008), and Regier and Kay (2009).
Arguably, Whorf considered knowledge of other languages crucial in
understanding the thoughts and beliefs of speakers of those languages,
and believed ‘that those who envision a future world speaking only one
tongue, whether English, German, Russian, or any other, hold a misguided
ideal and would do the evolution of the human mind the greatest disser-
vice’ (Whorf 1956, p. 244). This statement harmoniously aligns with the
basic tenet of the multi-competence framework, which places the bi/multi-
lingual person at centre stage in the investigation of language–thought
relationships. Recent approaches to this strand of bilingual cognition
research proceed from identifying specific linguistic differences in how a
particular concept is encoded, and then attempt to see how bilinguals
represent those conceptual domains in their cognition, and what variables
modulate the extent to which bilinguals display unique representations.
Cook (2002) views conceptual representation in bilinguals along an
integration continuum. At one end, language-specific concepts are com-
pletely separated and the bilingual draws from these separate

representations depending on the language context. At the other end of

the continuum, language-specific concepts are merged into unique repre-
sentations, sharing elements from both languages. Here I will describe in
detail two domains of investigation where I have been directly involved
(object and substance classification, and colour categorization), but
research is also being conducted in other areas too, like grammatical
gender (Kurinski and Sera 2011), time (Miles et al. 2011), and motion events
(Kersten et al. 2010).
A number of studies have focused on categorization of objects and
substances in noun class and classifier languages. Noun class languages
like English distinguish between count and mass nouns grammatically.
Count nouns refer to discrete entities that are marked for number (e.g.
apple – apples), while mass nouns refer to entities that are perceived as
inherently non-individuated (e.g. substances) and are not marked for num-
ber (e.g. water – *waters). In classifier languages like Japanese or Chinese,
there is no grammaticized count/mass distinction. Nouns in these lan-
guages refer to substances and are accompanied by numeral classifiers
(e.g. three small-piece-of apple [= 3 apples]). Speakers of the two types of
languages were shown to perform differently on a similarity judgment
task that required participants to match objects based on their common
shape or material. Speakers of English favoured shape and speakers of
Japanese favoured material, presumably because noun-class languages
draw speakers’ attention to discreteness of entities and classifier lan-
guages to material (Lucy 1992; Imai and Gentner 1997). Several studies
found that Japanese speakers of L2 English tended to make shape-based
similarity judgments significantly more than Japanese monolinguals, both
as a function of their length of stay in the L2 speaking country (Cook et al.
2006) as well as their increasing L2 proficiency (Athanasopoulos 2007). In
all of these studies, bilingual categorization behaviour was in between that
of the monolingual groups, suggesting that bilinguals draw on conceptual
elements from both of their languages while categorizing entities.
Another popular domain of investigation is colour categorization, prob-
ably because of the centrality of that domain in the language and thought
debate (see Regier and Kay 2009). Studies have focused on both semantic
representation and cognitive categorization. In studies of the first kind,
participants may be asked to name separate colour chips or to indicate the
best example (prototype) of particular colour terms on a colour chart.
Andrews (1994) and Athanasopoulos (2009), replicating and extending
earlier studies by Ervin (1961) and Caskey-Sirmons and Hickerson (1977),
gave such tasks to Russian and Greek speakers who were speakers of L2
English. Russian and Greek make an obligatory lexical distinction between
light and dark blue (ghalazio and ble in Greek, goluboj and sinij in Russian).
Results showed that L2 speakers shifted the prototypes of their native blue
categories towards the English blue category, primarily as a function of
length of stay in the L2 speaking country. Athanasopoulos (2009) further

demonstrated that the longer Greek speakers of L2 English had lived in the
UK, the more likely they were to judge light and dark blue colour chips as
more similar to each other, suggesting a weakening of the distinction
between ble and ghalazio in their cognitive representation of those cate-
gories. Still, the perception of the blue categories neither resembled that of
Greek monolinguals, nor that of English monolinguals, but was in
between the two monolingual patterns as revealed by measuring the
bilinguals’ underlying brain activity (Athanasopoulos et al. 2010).
The authors concluded that their findings can be explained through
Casasanto’s (2008) assumption that learning a language implies the
strengthening of associations between words, concepts, and their refer-
ents in the real world. When an individual comes into contact with a new
language and culture, new associations are formed, and, crucially, the
strength of the old associations may be readjusted, giving rise to merged
patterns of performance depending on the degree of conceptual
readjustment.
17.5 Discussion
The findings to date show that knowing and using more than one language
may have profound consequences for cognition. Generally, bilinguals
demonstrate enhanced cognitive abilities related to executive control,
and unique mental representations in domains like colour and objects.
Importantly, studies also show that these patterns are flexible, and may
change with increasing L2 expertise, or under specific experimental con-
ditions. It appears that several factors that pertain to language expertise
and knowledge itself, as well as factors that have more to do with socio-
cultural environment and context of bilingualism, may modulate cogni-
tive restructuring in the multi-competent mind. Below I discuss some of
those factors, both in relation to cognitive-general and to language-specific
effects of bilingualism on cognition.
17.5.1 General language proficiency

Athanasopoulos (2006) and Athanasopoulos and Kasai (2008) showed a
shift in cognitive patterns in L2 users with an advanced level of L2 profi-
ciency. L2 users with intermediate proficiency continued to exhibit L1
cognitive patterns. Cook et al. (2006) and Athanasopoulos (2009) on the
other hand did not find a significant effect of proficiency on the cognitive
patterns of their L2 users. One possibility that could account for the
differences between these studies is the way general proficiency was
measured. Athanasopoulos (2006) and Athanasopoulos and Kasai (2008)
used the Oxford Quick Placement Test (QPT 2001), which measures profi-
ciency in a range of aspects of the L2, such as vocabulary and cultural

knowledge, as well as grammar and syntax, and text comprehension. Cook

et al. (2006) and Athanasopoulos (2009) used the Nation vocabulary test
(Nation 1990), which only measures vocabulary in the L2. Therefore it
appears that when more sensitive measures of proficiency are used (such
as the QPT), proficiency effects are more likely to become apparent.
General L2 proficiency may also interact in interesting ways with execu-
tive control functions of bilinguals. Bialystok and Feng (2009) conducted a
memory task involving language interference (participants had to recall
words from lists that they had previously seen). The findings showed that
bilingual children made fewer intrusions from previous lists even though
they recalled the same number of words. For the adults, bilinguals recalled
more words than monolinguals when the scores were corrected for differ-
ences in vocabulary. In addition, there was a strong effect of vocabulary in
which higher vocabulary participants recalled more words irrespective of
language group. These results clearly showed that vocabulary size in an L2
plays an important role in verbal performance and memory. Interestingly,
Bialystok and Feng (2009) also found that bilinguals may compensate for
weaker language proficiency with their greater executive control to
achieve the same or better levels of performance in the memory task as
monolinguals.
17.5.2 Specific language proficiency

This refers to knowledge of the specific linguistic property under investi-
gation, and pertains more to language-specific effects of bilingualism on
cognition, since in those cases the assumption is that acquisition of spe-
cific linguistic properties in the L2 brings about changes in bilingual
cognition. Elicitation of such knowledge may take several forms, ranging
from free narratives to controlled conversations to picture descriptions
designed to elicit the desired form, to written tasks such as grammaticality
judgments and lexical retrieval tasks, focusing on specific linguistic con-
structs. For example, Athanasopoulos (2007) used a grammaticality judg-
ment task to measure participants’ knowledge of grammatical number
marking in English. This score was then combined with the score obtained
in the QPT to create a composite L2 proficiency index for each participant.
This score significantly predicted each L2 user’s degree of cognitive shift
towards L2-based patterns. Athanasopoulos (2009) compared lexical mem-
ory for colour terms with similarity judgments of colours and found that
availability of the specific colour terms in lexical memory correlated with
participants’ colour similarity judgments.
With regard to cognitive-general effects, several models of bilingualism
assume that greater cognitive control is required when the languages of
the bilingual are more similar to each other. Such interference effects
become apparent as proficiency in the L2 increases, and items from the
L2 compete for selection with items from the L1 during comprehension

and production. Interference effects may become stronger when target

items become more similar to items in the other language with respect to
orthography, phonology, or semantics (Dijkstra and van Heuven 1998). At
the phonological level the overlapping segments result in competition
between the highly similar words. Likewise, at the orthographic level a
number of lexical orthographic candidates are activated in parallel. In the
case of two languages with alphabetical writing systems (e.g. English and
Spanish), the number of activated orthographic candidates is much larger
than the number activated in bilinguals whose languages share no ortho-
graphy at all (e.g. English and Chinese). Greater cognitive effort is required
to attend and select among orthographically similar candidates than
among orthographically distant ones (Shook and Marian 2013). The impli-
cations of phonological and orthographic similarity/distance, and levels of
L2 proficiency and literacy in specific languages, scripts, and orthogra-
phies for non-linguistic executive function have yet to be systematically
explored.
17.5.3 Age of language acquisition

Development and mastery of a second language, as well as any advan-
tages in executive control through use of two or more languages, could
well depend on maturational constraints, either because of a biologi-
cally determined ‘sensitive’ period for language acquisition, or due to
gradual decline of general learning mechanisms throughout the life-
span. For this reason, the influence of this variable may not be directly
observable, but instead it may be a mediating variable in the relation-
ship between language proficiency and degree of cognitive restructur-
ing. For example, Athanasopoulos and Kasai (2008) found that both
specific L2 proficiency and age of L2 acquisition could predict L2 cog-
nitive shift, but at the same time the two predicting variables also
correlated with each other. In a subsequent analysis, the effect of
specific L2 proficiency on cognitive shift remained even when taking
into account age of L2 acquisition, while the effect of age of L2 acquisi-
tion was abolished when taking into account specific L2 proficiency.
On the other hand, Boroditsky (2001) found a strong effect of age of L2
acquisition, while length of exposure to English played no role
whatsoever.
In the domain of executive control, Luk et al. (2011) correlated perfor-
mance on the flanker task with English proficiency in early and late
bilingual adults (according to whether they became actively bilingual
before or after the age of ten years). Findings showed that early bilinguals
produced the smallest response time cost for incongruent trials (flanker
effect) with no difference between monolinguals and late bilinguals.
Moreover, across the whole sample of bilinguals, onset age of active
bilingualism was negatively correlated with English proficiency and

positively correlated with the flanker effect. The younger the age of onset
of bilingualism, the higher the L2 proficiency, and the lower the switching
cost in the flanker effect.
17.5.4 Frequency of language use

The degree to which an individual will use their second language may
often depend on the interactional context and the degree of immersion to
a specific community or country. Increasing the opportunity to use the
language due to these factors may in turn increase expertise in the parti-
cular language and may potentially provide the individual with target-like
examples of specific linguistic features. To date, very few studies have
directly attempted to link bilingual cognition to how frequently the bilin-
guals use or switch between their languages. In language-specific investi-
gations, the study by Athanasopoulos et al. (2011) explicitly measured the
role of frequency of use of the L2 whilst controlling for variables such as
proficiency, age of acquisition, and length of stay in the L2 speaking
country. The study did show a significant effect of frequency of language
use, such that the more frequently individuals used the L2, the more their
cognitive patterns shifted away from the L1 towards the L2.
In investigations of general effects of bilingualism on cognition, fre-
quency of usage is often operationalized as frequency of code-switching
or code mixing. Yim and Bialystok (2012) found that Cantonese–English
bilinguals who engaged in more conversational code-switching showed
advantages in verbal task switching relative to those bilinguals that code-
switched less frequently. However, performance in a non-verbal task-
switching task was not linked to degree of conversational code-switching.
17.5.5 Experimental context and task manipulation

In the context of evaluating perceptual distinctions, an L2 setting (e.g.
receiving experiment instructions in the L2 by a native speaker of the L2)
may promote a cognitive pattern that resembles that of monolingual
speakers of the L2. This is precisely what was found by Boroditsky et al.
(2002) in their study of action events. L2 users instructed in the L2 showed a
more L2-based cognitive pattern than L2 users tested in their L1. However,
the study also showed that L2 users who received instructions in the L1 did
not show a completely L1-based cognitive pattern, but were in between
monolingual speakers of both languages. Athanasopoulos (2007) also var-
ied the language of task instructions but found no significant difference.
Both L1-instructed and L2-instructed bilinguals displayed in-between cog-
nitive patterns, although some effects emerged when the bilingual groups
were compared to their monolingual peers. Those bilinguals instructed in
the L1 differed significantly from monolingual speakers of their L2,
whereas those instructed in the L2 did not.

Strong effects of language of experiment instruction were reported in

Kersten et al.’s (2010) study on motion event categorization, which
required Spanish–English bilinguals to sort novel, animated objects and
events into categories on the basis of manner of motion. Results showed
that when instructions were given in English, bilinguals’ behaviour moved
towards English patterns, and when instructed in Spanish their behaviour
moved towards Spanish patterns. However, this testing context effect was
modulated by age of L2 acquisition. Specifically, late bilinguals (age of
acquisition ≥ 5 years) were far more affected by the different testing
contexts than early bilinguals (age of acquisition < 5 years), who displayed
similar performance in both English and Spanish contexts.
In the area of executive control, Bialystok (2006) showed that bilinguals
were faster when the task required the most controlled attention to
achieve conflict resolution in the Simon task, highlighting the role of
bilingual experience as a modifier of performance. Interestingly, monitor-
ing effort to resolve conflicting information is maximized when the bilin-
gual individual is presented with an equal number of congruent and
incongruent trials in a mixed design (presenting both congruent and
incongruent stimuli as part of the same block), rather than an unequal
proportion of congruent and incongruent stimuli (Costa et al. 2009), or in
blocked designs (presenting congruent and incongruent stimuli in sepa-
rate blocks: see Costa et al. 2009 for a detailed meta-analysis).
17.5.6 Length of immersion in L2-speaking/bilingual community

From the perspective of language-specific investigations, living in the
country where the second language is spoken as a native language may
promote and induce in the individual an inclination, conscious or uncon-
scious, to emulate the linguistic, and crucially, the non-linguistic beha-
viour of the target-language community. Cook et al. (2006), for example,
found no between-group significant difference when they split their L2
users by general L2 proficiency, but they did find a significant difference
when L2 users were split by length of stay in the L2 country.
Athanasopoulos (2007) found a significant relationship between length
of stay and L2 cognitive shift, but the effect disappeared when L2 profi-
ciency was taken into account. Athanasopoulos (2009) found independent
effects of both lexical memory and length of stay on colour similarity
judgments, suggesting that both factors may influence cognition.
Bialystok and Barac (2012) examined the possible factors underpinning
the reported advantages in executive control for bilingual children. The
researchers studied children in the process of becoming bilingual by
attending immersion programmes. The findings showed that length of
time in the immersion programme was related to performance on execu-
tive control tasks, such that increase in the former correlated with more
efficient performance in the latter.

17.5.7 Context of acquisition/bilingualism

While length of bilingual immersion may lead to cognitive restructuring
in the mind of the L2 user, no study has systematically manipulated the
variable of acquisition context to compare, say, the effects of naturalistic
versus instructed learning on L2 user cognition, or the effects of different
types of sociocultural bilingual contexts. In language-specific investiga-
tions, studies have independently found effects of L2 acquisition both in
naturalistic (Athanasopoulos 2009) and in instructed L2 learning (Kurinski
and Sera 2011), but no study has compared the outcomes for cognitive
restructuring between the different ways bi/multilinguals learn their
languages.
In the cognitive-general arena, Costa et al. (2009) have raised the inter-
esting possibility that the reported advantages in executive control studies
employing mixed designs (see Section 17.5.5) are due to the fact that the
bilinguals studied live in a bilingual region where switching between
languages happens constantly and often unpredictably. Arguably, this
type of language switching resembles the mixed design used in the afore-
mentioned studies (Costa et al. 2009). Presumably, then, blocked designs
would favour more bilinguals who use their languages in specific contexts
rather than interchangeably, yet no study has directly attempted to corre-
late context of bilingualism with mixed or blocked experimental designs
in executive control tasks. It remains possible that executive control may
be enhanced by communicative behaviour that demands frequent mon-
itoring of the required language in societies where bilingualism is frequent
and established.
17.6 Conclusion
The current synthesis of the main findings in each strand of bilingual

cognition studies has made it clear that in the past two decades there has
been an extremely productive stream of empirical research on the effects of
bilingualism on cognition, both in terms of directly operationalizing lin-
guistic influences on non-verbal cognitive processes by means of linguistic,
behavioural, and more recently neurophysiological techniques, and in
terms of implementing tasks traditionally used in the attentional control
literature to investigate executive control processes and their neural corre-
lates in bilinguals. To date, cognitive-general and language-specific enqui-
ries have remained largely independent of one another. Being aware of
developments in both lines of enquiry is important for several reasons:
(i) Most prominent accounts of the mental architecture of bilingualism

assume that languages are equivalent at the conceptual level (e.g.
Kroll and Stewart 1994 and Costa 2005, but see also de Groot 1992),
and the task of the bilingual, once a concept is activated, is to select

the word in the intended language, while suppressing interference

from the word’s translation equivalent in the other language (Green
1998b). However, the need to re-examine the assumption of concep-
tual equivalence in the languages of bilinguals has recently been
highlighted (Pavlenko 2005; Jarvis and Pavlenko 2008) because firstly
the majority of words that may seem like translation equivalents
across languages do not have exactly the same conceptual representa-
tion, even when they refer to concrete entities (Ameel et al. 2005), and
secondly there is now a growing body of evidence from language-
specific investigations of bilingual cognition that shows systematic
cross-linguistic variation in conceptual representation of a range of
different domains of experience. Thus the basis of executive control
advantages in bilinguals (suppression of interference at the lexical
selection stage) needs to be revisited in light of the findings of lan-
guage-specific investigations that show that learning a new language
does not only imply mapping different words onto the same language-
independent concepts, but also creating new concepts and recalibrat-
ing existing concepts that are dissimilar to those of monolingual
speakers.
(ii) Most studies of language-specific effects of bilingualism on cognition
tend not to take into account the unique executive control skills of
bilinguals, yet taking these skills into account could potentially pro-
vide a breakthrough into one of the central debates in linguistic
relativity research, namely whether effects of language on cognition
are of a permanent or transient nature. Some studies show that
language has an impermeable grip on cognition, invading our think-
ing even when we are not consciously aware of it (e.g. Thierry et al.
2009; Athanasopoulos et al. 2009). Other studies show that cross-
linguistic differences in non-verbal similarity judgments disappear
in dual task paradigms where, concurrent to the cognitive task, parti-
cipants are asked to perform a verbal interference task to inhibit
inner speech (e.g. Winawer et al. 2007; Trueswell and Papafragou
2010). Thus verbal interference has been used as the hallmark of the
transient nature of the effects of language on cognition. However, this
effect has never been studied in bilinguals. Given that when bilin-
guals use one of their languages, the other language remains active
and fully accessible (Wu and Thierry 2010), it is entirely possible that
verbal interference may block out the cognitive behaviour associated
with the language of interference, such that the bilingual individual
may fall back on the other alternative behaviour (i.e. that of the non-
interfered language). This would reveal the full extent of cognitive
malleability in the bilingual, and by extension, human mind.
(iii) The majority of linguistic relativity research to date has not paid
due attention to the fact that the majority of humans use more
than one language in their everyday functions, and may have

different cognitive abilities and mental representations from

monolinguals. This may have serious scientific implications. For
example, a key study that was cited as evidence for universality
and against relativity showed that native speakers of 20 different
languages displayed English-like category prototypes (Berlin and
Kay 1969). The authors admitted that the speakers in their sample
were for the most part US immigrants with various levels of accul-
turation and English proficiency, yet they ‘find it hard to believe
that English could so consistently influence the placement of the
foci in these diverse languages’ (Berlin and Kay 1969, p. 12).
Subsequent studies found profound effects of bilingualism on pro-
totype identification and colour categorization (Caskey-Sirmons
and Hickerson 1977; Athanasopoulos 2009), and thus cast serious
doubt on the idea that using another language cannot affect think-
ing (subsequent studies by Paul Kay and his team controlled more
rigorously for linguistic and cultural differences in their samples:
see Regier and Kay 2009, for a review). In the domain of motion
events, a series of ground-breaking studies by Papafragou and
colleagues (e.g. Papafragou and Selimis 2010) have shown mixed
results regarding the influence of language on event cognition, but
have given little consideration to the possibility that their partici-
pants may be influenced by knowledge of another language. Yet
studies show that bilinguals are indeed influenced by their L2
when judging the similarity between different types of motion
events (Bylund and Athanasopoulos 2014), and that they may
already have different conceptualizations for motion events even
at intermediate stages of L2 proficiency (Gullberg 2011).
To conclude, we have now reached a critical point in bilingual

cognition studies, where the two aforementioned approaches are pro-
ceeding with their own research agendas without cross-fertilization of
ideas and findings. To date no studies have yet established what role
bilingualism may play in many key findings in the linguistic relativity
literature, and conversely, most accounts of the source of executive
control in bilinguals assume that languages are equivalent at the con-
ceptual level. This necessarily limits our view of bilingual cognition,
and by extension human cognition, since multilinguals may well
represent the largest portion of humanity (Cook 2002). Now is the
time to pursue a concerted, multidisciplinary approach to build on
the exciting new developments in cognitive-general and language-spe-
cific investigations and create research agendas that will be more
sensitive and at least cognizant of developments in both areas of
bilingual cognition studies. The framework of multi-competence read-
ily brings together these hitherto separate strands of research into the
bilingual mind.

References
Abutalebi, J. and Green, D. W. 2008. ‘Control mechanisms in bilingual

language production: neural evidence from language switching studies’,
Language and Cognitive Processes 24, 557–582.
Ameel, E., Storms, G., Malt, B. and Sloman, S. 2005. ‘How bilinguals solve
the naming problem’, Journal of Memory and Language 53, 60–80.
Andrews, D. 1994. ‘The Russian color categories sinij and goluboj: an experi-
mental analysis of their interpretation in the standard and emigré lan-
guages’, Journal of Slavic Linguistics 2, 9–28.
9, 89–96.
Athanasopoulos, P. 2007. ‘Interaction between grammatical categories
and cognition in bilinguals: the role of proficiency, cultural immer-
sion, and language of instruction’, Language and Cognitive Processes 22,
689–699.
Athanasopoulos, P. 2009. ‘Cognitive representation of colour in bilin-
guals: the case of Greek blues’, Bilingualism: Language and Cognition 12,
83–95.
Athanasopoulos, P. 2011. ‘Cognitive restructuring in bilingualism’, in
A. Pavlenko (ed.), pp. 29–65.
Japanese blues’, Bilingualism: Language and Cognition 14, 9–17.
Athanasopoulos, P., Dering, B., Wiggett, A., Kuipers, J. R. and Thierry, G.
2010. ‘Perceptual shift in bilingualism: brain potentials reveal plasticity
in pre-attentive colour perception’, Cognition 116, 437–443.
Athanasopoulos, P., Wiggett, A., Dering, B., Kuipers, J. and Thierry, G. 2009.
‘The Whorfian Mind: electrophysiological evidence that language
shapes perception’, Communicative and Integrative Biology 2, 332–334.
Berlin, B. and Kay, P. 1969. Basic Color Terms: Their Universality and Evolution.
Berkeley: University of California Press.
Bialystok, E. 2006. ‘Effect of bilingualism and computer video game experi-
ence on the Simon task’, Canadian Journal of Experimental Psychology 60,
68–79.
Bialystok, E. 2009. ‘Bilingualism: the good, the bad, and the indifferent’,
Bialystok, E. and Barac, R. 2012. ‘Emerging bilingualism: dissociating
advantages for metalinguistic awareness and executive control’,

Bialystok, E., Craik, F. I. M. and Freedman, M. 2007. ‘Bilingualism as a

protection against the onset of symptoms of dementia’,
Neuropsychologia 45, 459–464.
Bialystok, E., Craik, F. I. M., Klein, R. and Viswanathan, M. 2004.
‘Bilingualism, aging, and cognitive control: evidence from the Simon
task’, Psychology and Aging 19, 290–303.
Bialystok, E., Craik, F. I. M. and Luk, G. 2008. ‘Cognitive control and lexical
access in younger and older bilinguals’, Journal of Experimental Psychology:
Learning, Memory and Cognition 34, 859–873.
Bialystok, E. and Feng, X. 2009. ‘Language proficiency and executive con-
trol in proactive interference: evidence from monolingual and bilingual
children and adults’, Brain and Language 109, 93–100.
Boroditsky, L. 2001. ‘Does language shape thought? English and Mandarin
speakers’ conceptions of time’, Cognitive Psychology 43, 1–22.
Boroditsky, L., Ham, W. and Ramscar, M. 2002. ‘What is universal in
event perception? Comparing English & Indonesian speakers’, in
W. D. Gray and C. D. Schunn (eds), Proceedings of the 24th Annual
Conference of the Cognitive Science Society, pp. 136–141. Lawrence
Erlbaum.
Bylund, E. and Athanasopoulos, P. 2014. ‘Language and thought in a multi-
lingual context: the case of isiXhosa’, Bilingualism: Language and Cognition
17, 431–441.
Casasanto, D. 2008. ‘Who’s afraid of the Big Bad Whorf? Cross-linguistic
differences in temporal language and thought’, Language Learning 58,
63–79.
Caskey-Sirmons, L. and Hickerson, N. 1977. ‘Semantic shift and bilingual-
ism: variation in the color terms of five languages’, Anthropological
Cook, V. J 1992. ‘Evidence for multi-competence’, Language Learning 42,
557–591.
Cook, V. J 1999. ‘Going beyond the native speaker in language teaching’,
TESOL Quarterly 33, 185–209.
Cook, V. J. 2002. ‘Background to the L2 user’, in V. J. Cook (ed.), Portraits of
Cook, V. J 2003. ‘The changing L1 in the L2 user’s mind’, in V. J. Cook (ed.),
Effects of the Second Language on the First, pp. 1–18. Clevedon: Multilingual
Matters.
Cook, V. J. and Bassetti, B. (eds) 2011. Language and Bilingual Cognition.
New York: Psychology Press.
Cook, V. J, Bassetti, B., Kasai, C., Sasaki, M. and Takahashi, J. 2006.
‘Do bilinguals have different concepts? The case of shape and material
in Japanese L2 users of English’, International Journal of Bilingualism 10,
137–152.

Costa, A. 2005. ‘Lexical access in bilingual production’, in J. F. Kroll and

A. M. B. de Groot (eds), Handbook of Bilingualism: Psycholinguistic Approaches,
pp. 308–325. New York: Oxford University Press.
Costa, A., Hernández, M., Costa-Faidella, J. and Sebastián-Gallés, N. 2009.
‘On the bilingual advantage in conflict processing: now you see it, now
you don’t’, Cognition 113, 135–149.
Cummins, J. 1976. ‘The influence of bilingualism on cognitive growth: a
synthesis of research findings and explanatory hypotheses’, Working
Papers on Bilingualism 9, 1–43.
Phonology, Morphology, and Meaning, pp. 389–412. Amsterdam: North-
Holland.
Dijkstra, A. (Ton) and van Heuven, W. J. B. 1998. ‘The BIA model and
bilingual word recognition’, in J. Grainger and A. M. Jacobs (eds),
Localist Connectionist Approaches to Human Cognition, pp. 189–225.
Emmorey, K., Luk, G., Pyers, J. E. and Bialystok, E. 2008. ‘The source of
enhanced cognitive control in bilinguals’, Psychological Science 19, 1201–
1206.
Ervin, S. 1961. ‘Semantic shift in bilingualism’, American Journal of Psychology
74, 233–241.
Ervin, S. 1964. ‘Language and TAT content in bilinguals’, Journal of Abnormal
and Social Psychology 68, 500–507.
Gentner, D. and Goldin-Meadow, S. (eds) 2003. Language in Mind: Advances in
the Study of Language and Thought. Cambridge, MA: MIT Press.
Green, D. W. 1998a. ‘Bilingualism and thought’, Psychologica Belgica 38,
253–278.
Green, D. W. 1998b. ‘Mental control of the bilingual lexico-semantic sys-
tem’, Bilingualism: Language and Cognition 1, 67–81.
Grosjean, F. 1982. Life with Two Languages: An Introduction to Bilingualism.
Cambridge University Press.
Grosjean, F. 1992. ‘Another view of bilingualism’, in R. Harris (ed.),
Cognitive Processing in Bilinguals, pp. 51–62. Amsterdam: North-Holland.
Gullberg, M. 2011. ‘Thinking, speaking, and gesturing about motion in
more than one language’, in A. Pavlenko (ed.), pp. 143–169.
Ianco-Worrall, A. D. 1972. ‘Bilingualism and cognitive development’, Child
Imai, M. and Gentner, D. 1997. ‘A crosslinguistic study of early word
169–200.

Jarvis, S. (ed.) 2011. Crosslinguistic Influence in Bilinguals’ Concepts and

Conceptualizations, special issue of Bilingualism: Language and Cognition 14.
Kersten, A. W., Meissner, C. A., Lechuga, J., Schwartz, B. L., Albrechtsen, J. S.
and Iglesias, A. 2010. ‘English speakers attend more strongly than
Spanish speakers to manner of motion when classifying novel objects
and events‘, Journal of Experimental Psychology: General 139, 638–653.
picture naming: evidence for asymmetric connections between bilin-
gual memory representations’, Journal of Memory and Language 33,
149–174.
Kurinski, E. and Sera, M. D. 2011. ‘Does learning Spanish grammatical
gender change English-speaking adults’ categorization of inanimate
objects?’, Bilingualism: Language and Cognition, 14, 203–220.
Laurén, U. 1991. ‘A creativity index for studying the free written produc-
tion of bilinguals’, International Journal of Applied Linguistics 1, 198–208.
Levinson, S. C. 1996. ‘Relativity in spatial conception and description’, in
J. J. Gumperz and S. C. Levinson (eds), Rethinking Linguistic Relativity,
Lucy, J. A. 1992. Grammatical Categories and Cognition: A Case Study of the
Linguistic Relativity Hypothesis. Cambridge University Press.
Luk, G., de Sa, E. and Bialystok, E. 2011. ‘Is there a relation between onset
age of bilingualism and enhancement of cognitive control?’, Bilingualism:
Miles, L. K., Tan, L., Noble, G. D., Lumsden, J. and Macrae, C. N. 2011. ‘Can a
mind have two time lines? Exploring space–time mapping in Mandarin
and English speakers’, Psychological Bulletin and Review 18, 598–604.
Nation, P. 1990. Teaching and Learning Vocabulary. New York: Newbury
House/Harper Row.
Papafragou, A. and Selimis, S. 2010. ‘Event categorisation and language: a
cross-linguistic study of motion’, Language and Cognitive Processes 25,
224–260.
Pavlenko, A. 1999. ‘New approaches to concepts in bilingual memory‘,
Groot (eds), Handbook of Bilingualism: Psycholinguistic Approaches, pp. 433–
453. Oxford University Press.
Pavlenko, A. 2011. Thinking and Speaking in Two Languages. Bristol:
gence’, Psychological Monographs 76, 1–23.
Quick Placement Test. 2001. Oxford University Press.
The Radein Initiative: Some Directions in Research on Multilingualism.
http://groupcien.uab.es/documents/Radein_initiative_v2.0.pdf

Regier, T. and Kay, P. 2009. ‘Language, thought and color: Whorf was half
right’, Trends in Cognitive Sciences 13, 439–446.
Roberson, D. 2005. ‘Color categories are culturally diverse in cognition as
well as in language’, Cross-Cultural Research 39, 56–71.
Saer, D. J. 1923. ‘The effects of bilingualism on intelligence’, British Journal
of Psychology 14, 25–38.
Saer, D. J., Smith, F. and Hughes, J. 1924. The Bilingual Problem: A Study Based
upon Experiments and Observations in Wales. Aberystwyth: University
College of Wales.
Shook, A. and Marian, V. 2013. ‘The Bilingual Language Interaction
Network for comprehension of speech‘, Bilingualism: Language and
Thierry, G. and Wu, Y. J. 2007. ‘Brain potentials reveal unconscious transla-
tion during foreign language comprehension’, Proceedings of the National
Academy of Sciences 104, 12530–12535.
Thierry, G., Athanasopoulos, P., Wiggett, A., Dering, B. and Kuipers, J. 2009.
‘Unconscious effects of language-specific terminology on pre-attentive
color perception’, Proceedings of the National Academy of Sciences 106, 4567–
4570.
Trueswell, J. C. and Papafragou, A. 2010. ‘Perceiving and remembering
events cross linguistically: evidence from dual-task paradigms’, Journal
of Memory and Language 63, 64–82.
von Humboldt, W. 1963. Humanist without Portfolio: An Anthology of the
Writings of Wilhelm von Humboldt. Detroit, MI: Wayne State University
Press.
Whorf, B. L. 1956. Language, Thought, and Reality: Selected Writings of Benjamin
Lee Whorf (J. B. Carroll, ed.). Cambridge, MA: MIT Press.
Winawer, J., Witthoft, N., Frank, M. C., Wu, L., Wade, A. R. and Boroditsky,
L. 2007. ‘Russian blues reveal effects of language on color discrimina-
tion’, Proceedings of the National Academy of Sciences 104, 7780–7785.
hear Chinese’, Journal of Neuroscience 30, 7646–7651.
Yim, O. and Bialystok, E. 2012. ‘Degree of conversational code-switching
enhances verbal task switching in Cantonese–English bilinguals’,

18
Space, motion and
thinking for language
Anna Ewert
Although the term multi-competence was originally defined in reaction to

inadequacies of the then prevalent model of language as ‘the compound
state of a mind with two grammars’ (Cook 1991, p. 112), it was soon
redefined as ‘knowledge of two or more languages in one mind’ (Cook
2003, p. 2) in recognition of the widespread cognitive consequences of
knowing more than one language (see Cook’s 1992 review, as well as
Athanasopoulos, Chapter 17, this volume). Multi-competence research
into language and cognition started with two studies on number cognition
in L2 users (Cook et al. 2006; Athanasopoulos 2006). Both adopted the
Whorfian perspective to account for the nature of the influence of lan-
guage on cognition. Research on motion cognition followed (e.g. Brown
2008; Brown and Gullberg 2008), adopting Slobin’s thinking for speaking
perspective. Yet another theoretical perspective, embodied cognition, has
been adopted in the more recent Tomczak and Ewert (2015) study. This
variety creates some kind of a tension, suggesting that perhaps one of
these perspectives is better than others, or more appropriate for studying
the multi-competent mind. It will be argued below that these three theo-
retical perspectives deal with different constructs and inform us about
somewhat different aspects of human cognition.
The present chapter compares the three different theoretical perspec-
tives adopted in research on motion and spatial cognition and reviews
relevant research with both monolingual participants and second
language users. The first of these perspectives, the linguistic relativity
hypothesis, raised the most controversies in the past. Paradoxically, with
new research in the area of categorical perception, this perspective seems
to be more vibrant nowadays than ever before. The thinking for speaking
perspective is the one most often adopted in cross-linguistic and second
This research has been financed in part by grant N N104 380040 from the Polish Ministry of Science and Higher
Education and the National Science Centre to Anna Ewert.

Space, motion and thinking for language 377
language research in the domain of motion. New eye tracking research

shows that its explanatory potential has not been exhausted yet. The
embodiment perspective offers an entirely new view of human cognition,
but its potential for cross-linguistic and second language research has not
been realised yet.
The discussion focuses on differences in the theoretical constructs used
and explanations for the locus and the mechanism of the influence of
language on cognition proposed within each of these perspectives. Cross-
linguistic and second language research is reviewed where appropriate.
It is concluded that each of these perspectives is currently vibrant and
promises to yield interesting results, contributing to better understanding
of human space and motion cognition.
18.1 Linguistic relativity
Although Whorf had predecessors, he remains the key figure in the cur-
rent language and thought debate, focusing on questions such as: Does
language shape perception? If indeed it does, does that mean that speakers
of a particular language are not able to make perceptual distinctions that
their language does not encode? Are different aspects of cognition influ-
enced by language to the same extent?
Over the years, Whorf’s relativity hypothesis raised numerous contro-
versies, stemming in part from misunderstanding, partly from misinter-
pretation, the most fundamental attacks being launched by the so-called
universalists, who assumed that conceptual representations are either
inborn and immutable or based entirely on physical experience. Instead
of summarising that debate, the arguments below will focus on what
Whorf said and how it is relevant for present-day research on space and
motion (however, see Pavlenko 2005, 2011 or Casasanto 2008 for a sum-
mary of the most important arguments not discussed here).
18.1.1 Definition, locus of the language effect and

operationalisation
In Whorf’s words:
The linguistic relativity principle . . . means . . . that users of markedly

different grammars are pointed by their grammars toward different
types of observations and different evaluations of externally similar acts
of observation, and hence are not equivalent as observers but must arrive
at somewhat different views of the world. (Whorf 1940a/1956, p. 221)
While Whorf (1941/1956, p. 158) carefully explains that by ‘grammar’ he

actually means ‘fashions of speaking’, i.e. ‘lexical, morphological, syntac-
tic, and otherwise systemically diverse means coordinated in a certain

378 ANNA EWERT
frame of consistency’, ‘different views of the world’ resonate with echoes

of discussions of time long past. Since, unlike his predecessors, Whorf as a
matter of fact does not write about different world views but, instead,
gives examples of the effect of language on people’s behaviour, like in the
famous example of empty gasoline drums that were considered to be safe
to smoke cigarettes nearby, so it can be safely assumed that in present-day
terms the definition refers to the effect of language on mental representa-
tion of reality and its behavioural consequences. Hence, in present-day
psycholinguistic research, the Whorfian effect on cognition is frequently
operationalised as behavioural differences between speakers of
different languages on non-linguistic tasks (but see Lucy 1997 for an over-
view of other approaches). If such differences can be found, they are
interpreted as pointing to differences in conceptual organisation that
affect categorical perception. This line of reasoning can also be found in
Pavlenko’s (2005, p. 435) definition of concepts as ‘mental representations
that affect individuals’ immediate perception, attention, and recall and
allow members of specific language and culture groups to conduct identi-
fication, comprehension, inferencing and categorisation along similar
lines’.
Since the concept is the locus of language effect on cognition, it is
essential to look at how Whorf defines concepts. Whorf’s definition of a
concept is surprisingly modern:
The world is presented in a kaleidoscopic flux of impressions which has to

be organized by our minds – and this means largely by the linguistic
systems in our minds. We cut nature up, organize it into concepts, and
ascribe significances as we do, largely because we are parties to an agree-
ment to organize it in this way – an agreement that holds throughout our
speech community and is codified in patterns of our language.
(Whorf 1940b/1956, pp. 213–214)
Translating this into present-day terminology, we can think of a concept

as a collection of exemplars on which language imposes some kind of
internal organisation. The proposed mechanism allowing language to
shape mental categories is purely associationist. Whorf (1941/1956,
p. 147) calls it habitual thought and explains that linguistic patterns have
analogical and suggestive influence on how people think. Like many other
linguists of his time (e.g. Bloomfield 1933), he saw language as a habit in a
very much Pavlovian sense of stimulus–response associations; he speaks
of ‘linguistic conditioning of behaviour’ (Whorf 1941/1956, p. 135) and
‘behavioural compulsiveness’ of language (Whorf 1941/1956, p. 137).
Casasanto (2008) explicates this learning mechanism in more contempor-
ary terms as an instance of Hebbian learning through which certain
conceptual mappings are strengthened by repeated use of specific linguis-
tic constructions. Hence, conceptual representations of reality are both
formed through physical experience and shaped by language.

Having said that in Whorf’s view language conditions thought and

behaviour, several qualifications must be made. As Whorf’s own definition
of linguistic relativity includes an inherent operationalisation, this led to
the formulation of the so-called strong version of linguistic relativity,
known also as linguistic determinism, maintaining that language deter-
mines thought. As pointed out by Pavlenko (2011, p. 19), this formulation
should rather be called the Brown–Lenneberg hypothesis, as it is an
oversimplification of Whorf’s original statement. Whorf was interested
in categories of thought and his methods were ethnographic, while Brown
and Lenneberg (1954) were interested in the effect of language on cogni-
tive processes and sought to provide experimental verification of the
Whorfian hypothesis. The experiment they conducted required a direc-
tional hypothesis and a design that would produce measurable outcomes.
18.1.2 Categorical perception: colour versus space and motion

Brown and Lenneberg (1954) were interested in the effect of linguistic
categorisation on recognition memory for colour. The study was impor-
tant, because it provided the first empirical verification of the Whorfian
hypothesis, confirming that colours that would represent prototypical
exemplars of a corresponding linguistic category were remembered better
than shades of colour that were not easily classified as typical examples of
a particular category. It also inspired a lot of subsequent research into
colour cognition. Colour turned out to be a good test case for the Whorfian
hypothesis, since the stimuli could be relatively easily controlled. Selected
findings from this research will be presented briefly here, since they are
relevant for all domains of human cognition, including space and motion.
Findings in the domain of colour perception do generalise to other
domains of cognition (Gilbert, Regier, Kay and Ivry 2008), nevertheless
percepts that are more complex and involve more abstract representations
may be processed somewhat differently from colour stimuli.
Many researchers’ position on the Whorfian question has been for years
determined by Berlin and Kay (1969), which is still sometimes called upon
as an argument for the universalist position that language does not influ-
ence basic cognition. However, what this universalist tradition research
really demonstrates is simply that different languages do not divide the
colour spectrum in a completely arbitrary way (see Kay, to appear, for a
recent summary). This means that certain features of human perception
influence the way very different languages categorise the observed reality.
In other words, this research informs us about the influence of perception
on linguistic categorisation, but not about the influence of language on
cognition. To see how linguistic categories influence colour cognition, we
must look into the results of empirical psycholinguistic research.
While colour has been the test-bed for the linguistic relativity hypoth-
esis, research in the domain of space and motion followed. Levinson (1996,

380 ANNA EWERT
2003) distinguished three frames of reference that speakers of different

languages employ to describe the same spatial scene: intrinsic (‘He is in
front of the house’), relative (‘He is to the left of the house’) and absolute (‘He
is north of the house’). His research with Tzeltal and Danish speakers has
shown that the preferred frame of reference affected participants’
recognition and recall memory for spatial arrays and the inferences they
made with regard to the position of objects in space. Boroditsky, Ham and
Ramscar (2002) examined how grammatical tense affects perception of
human action and recognition memory using pictures portraying same
and different actors performing different actions. Their results show that
speakers of English perceive same tense actions by different actors to be
more similar than speakers of Indonesian, which does not have obligatory
tense marking on the verbs. Indonesian–English bilinguals performed
differently on these tasks depending on the language of testing. Tested in
English, they performed like monolingual speakers of English. Tested in
Indonesian, they performed like monolingual Indonesians on the recogni-
tion task, but they rated same actor, different tense pairs as less similar
than monolingual Indonesians did. In a similar experiment, Czechowska
and Ewert (2011) found that speakers of L1 Polish under the influence of
L2 English get desensitised to manner information and pay more attention
to the path of motion while rating the similarity of pictures presenting
simple motion events.
Inasmuch as Levinson’s (1996, 2003) and Boroditsky et al.’s (2002)
research provided initial evidence for linguistic relativity in the domain
of space and motion, a number of studies (Gennari, Sloman, Malt and Fitch
2002; Papafragou, Massey and Gleitman 2002) were published at around
the same time demonstrating lack of the expected relativity effect. For
example, Gennari et al. (2002) found no relativity effect in non-verbal tasks
in any of the experimental conditions but one, concluding that conceptual
and linguistic representations of motion are dissociable.
There are three possibilities to be accounted for: (i) there is no linguistic
relativity effect in the areas under study, since these are subject to some
universals in human perception, (ii) the instruments employed in these
studies were not sensitive enough to detect a relativity effect, (iii) there is
some confound that needs to be accounted for in further space and motion
research.
Kay’s (e.g. Cook, Kay and Regier 2005; Kay, to appear) research into
universals in colour naming informs us that there are certain areas
where relativity effects are not to be expected. Universals in talking
about space and motion have not really been investigated so far, and the
lack of language effect on behaviour demonstrated in some studies of
motion cognition cannot really be interpreted as showing that we have
identified an area in which the Whorfian hypothesis does not hold in
principle. Levelt (1996) notes that the frames of reference of the kind
described by Levinson might suggest that there are some universals in

the perception of space, since there are relatively few such frames, but
more systematic evidence of the kind that is available for colour categor-
isation across different languages (Cook, Kay and Regier 2005) is lacking
for how different languages categorise different aspects of space and
motion. A study by Malt et al. (2008) who examined how speakers of
English, Japanese, Spanish and Dutch name different gaits suggests that
at least some features of motion may be perceptually universal and inde-
pendent of language. This study shows that, although the respective lan-
guages use different verbs to describe different instances of human
locomotion, they all distinguish between instances of walking and run-
ning, and speakers of those languages perceive biomechanical discontinu-
ity between walking and running, which suggests that there might be
some universals in how humans perceive physical motion.
Another contested issue in Whorfian research has been the permanence
of the language effect. The question here is whether the effect of language
on cognition is permanent, or perhaps it is present only at the moment of
speaking. Kay and Kempton (1984) hypothesise that the Whorfian effect in
colour cognition is due to the participants’ use of the naming strategy, that
is tacitly, subvocally or even completely unconsciously naming the sti-
muli. To demonstrate this point they used a modified version of the triad
task in which the participants saw only two stimuli at a time, i.e. the target
and an alternate that was to be compared to the target. Additionally, each
of the alternates was named by the experimenter. Each of the two alter-
nates was equally distant from the target and the participants’ task was to
decide which of the two alternates was less like the target. Kay and
Kempton (1984) argue that this way of presentation of the stimuli reduces
the opportunity for naming the target by the participant. When the experi-
ment was administered in this fashion, the Whorfian effect indeed dis-
appeared. More evidence for the naming strategy comes from a set of
experiments conducted by Roberson and Davidoff (2000), who examined
the effect of verbal interference on categorical perception. In these experi-
ments the participants had to read out a word or count out loud while
committing the test stimuli to memory. The idea here is that since the
areas in the brain responsible for speech production are occupied by a
concurrent task, the participants will be prevented from naming the
colour stimuli, which indeed affected their categorical perception of
colour.
Similar results to the ones described in colour cognition have been
obtained in a number of studies dealing with motion cognition. Gennari,
Sloman, Malt and Fitch (2002) show that Spanish–English bilinguals
demonstrate language-specific behaviours in a similarity judgement task
with video clips of motion presented in triads, but only when they had
been asked to name the events prior to the non-verbal similarity judge-
ment part of the experiment. The language effect disappears when verbal
interference has been introduced, so that the participants were prevented

382 ANNA EWERT
from naming the scenes tacitly. In an eye-tracking study of motion cogni-

tion with native speakers of English and Greek (Papafragou, Hulbert and
Trueswell 2008), participants displayed language-specific gaze behaviour
while watching video clips of motion scenes, focusing their eyes on event
components that are typically encoded in their native language, but only
when they were asked to describe the scenes and not when asked to
inspect them freely. Trueswell and Papafragou (2010) demonstrate that
the effect of language on gaze behaviours is eliminated when verbal
interference is introduced. Athanasopoulos and Bylund (2013) also used
video clips of motion events presented in triads in a similarity judgement
task with speakers of Swedish and English. While the two groups differed
on this task when they had to commit the events to memory in order to
compare them, this difference disappeared when verbal interference was
introduced. These results suggest that the effect of language on motion
cognition, as is the case with colour cognition, is not permanent and the
Whorfian effect on motion perception can be demonstrated only when
language is available online.
On the other hand, Flecken, von Stutterheim and Carroll (2014) demon-
strate an effect of the participants’ language on gaze allocation patterns
while watching motion scenes in an auditory interference task; however,
the auditory stimuli they use are non-linguistic, so they do not rule out
subvocal naming of the visual stimuli by the participants. Interestingly,
Trueswell and Papafragou (2010, Experiment 1) also had a non-linguistic
interference condition, in which the participants had to tap out a rhythm
they heard while watching a video clip, and language did affect gaze
patterns in this condition. Trueswell and Papafragou (2010) argue that
this effect is due to an increased cognitive load of this condition, since
the additional task makes committing details of the event to memory
difficult, so the increased attentional demands of the task might incline
the participants to resort to linguistic encoding. While Flecken et al. (2014)
argue that their auditory interference results point to an early effect of
language on basic perception in a task in which language is not explicitly
involved, it seems that additional evidence is needed to resolve the
question of what the non-verbal interference results inform us about,
and how deep the influence of language on cognition is.
Thierry et al. (2009) were the first to demonstrate a difference in colour
perception between English and Greek native speakers in an event-related
brain potentials (ERP) study. They showed that speakers of Greek, which
has two categorically different blue colours, reacted more strongly to a
change in luminance while looking at blue stimuli than speakers of
English in an oddball shape discrimination task that focuses the partici-
pant’s attention on shape and not on the colour of the stimuli. This task
elicits a negativity at 200 ms after stimulus presentation (N200), the
so-called visual mismatch negativity (vMMN), and is considered to examine
differences in categorical perception at an automatic and pre-attentive

stage of processing. In another ERP study, Mo et al. (2011) found a

Whorfian effect in colour perception at around 200 ms post-stimulus,
but only when the stimuli were presented in the right visual field, which
means that categorical perception of colour is lateralised in the left
hemisphere of the brain. These results are considered to be indicative of
pre-attentional processing, but it does not mean that this processing takes
place at a pre-linguistic stage. Research shows that language is available
online as early as 200 ms (Costa et al. 2009), which again means that the
difference between speakers of different languages is related to simulta-
neous linguistic processing, i.e. conceptual processing taking place when
language is already available online, even though the language user may
not be consciously aware of it.
Moreover, the Thierry et al. (2009) experiment also showed a difference
between Greek and English speakers at an earlier stage of visual proces-
sing (P100), i.e. as early as 100 ms after stimulus presentation. This differ-
ence is quite difficult to interpret since there is only a handful of studies
analysing visual perception at this early stage; however, it seems to be
related to differences in categorical processing. Maier, Glage, Hohlfeld
and Abdel Rahman (2014) demonstrate that newly acquired verbal cate-
gories modulate the P100 component when stimuli are presented in the
right visual field, thus confirming left-lateralisation of visual categorical
perception. Since theirs was a linguistic category learning experiment, this
finding confirms that linguistic categorisation can influence low-level
perceptual processes.
18.1.3 So what are the space and motion categories?

Whorf (1941/1956, pp. 158–159) was very well aware that while language
affects the concepts people form, the relation between language, thought
and culture is a much more complex one. First, he was very well aware that
while some concepts, e.g. ‘time’ or ‘matter’, are more amenable to the
influence of language, others, like ‘space’, may be shaped more on the
basis of perception and physical experience than under the influence of
language. As he says, concepts are only ‘in part conditioned by the struc-
ture of particular languages’ (Whorf 1941/1956, p. 138), this relationship
being causal and linear. Second, linguistic concepts are only part of a
whole system of thought that is in complex ways connected to culture,
but as Whorf (1941/1956, p. 159) says, ‘[t]here are connections but not
correlations or diagnostic correspondences between cultural norms and
linguistic patterns’. As Nick Yee (1999) noticed in an unpublished paper, as
a chemical engineer by training, Whorf was probably at least to some
degree familiar with the fledgling science of quantum relativity and quan-
tum mechanics. He saw language, thought and culture as three separate
but interacting entities forming what we would call today a dynamic
system. Notwithstanding the non-linear and unpredictable interactions

384 ANNA EWERT
between these three, he had no doubt that language affects concepts as

mental categories.
While Whorfian-tradition research has traditionally revolved around
categorical perception, the influence of culture on conceptual
structures has been somewhat neglected. One well-known case, demon-
strating that such an influence exists, is the so-called Müller-Lyer illu-
sion in length estimation. The work by Segall, Campbell and Herskovits
(1966) demonstrates that this illusion is not a universal feature of
human visual perception and members of different societies are differ-
entially susceptible to this illusion, with representatives of some non-
industrialised societies, like the San foragers of the Kalahari, being
completely invulnerable to it.
Another example is spatial judgements. While members of western
cultures make absolute spatial judgements in visual tasks, disregarding
the background while estimating the size of an object, East Asians’ judge-
ments are always relative to the context in which the object is embedded.
Hedden et al. (2008) conducted a functional magnetic resonance imaging
(fMRI) study to find out if these two kinds of spatial judgements involve
different regions of the brain. When instructed to perform both types of
judgement, i.e. culturally preferred as well as culturally unpreferred, East
Asian and American participants did not differ significantly in accuracy or
response times. Imaging results did not identify differences in brain
regions responsible for visual or spatial processing either, but identified
increased activation in regions linked to executive control while making a
culturally unpreferred judgement. These findings are particularly relevant
for second language acquisition and bilingualism research on the acquisi-
tion of spatial concepts, as they indicate that such mental structures can be
acquired in an L2 to the point where performance is indistinguishable
from native speakers of the language, but the L2 user might need to pay
more attention while making culturally appropriate spatial judgements.
Cultural differences such as the ones described above do not necessarily
correspond to any specific differences in language, unless we insist that
the word for ‘bigger’ means something else in those different languages,
which raises an issue of correspondence between concepts and words.
The Whorfian hypothesis has been traditionally equated with the effect
of language on conceptual representations of reality. Thus, language
would affect categorical perception. The problem is relatively simple
when this reality can be operationalised as a continuum which can be
fairly arbitrarily divided into sections that can receive different linguistic
labels, as is the case with the colour continuum. In the case of colour
concepts, there is a one-to-one equivalence between concepts and the
words that represent them. A concept underlying the word blue can be
thought of as some part of a three-dimensional continuum containing
prototypical, i.e. the most typical shades of blue, as well as some borderline
exemplars, such as for example a greenish kind of blue. This

operationalisation, however, runs into some problems with different con-

ceptual structures that are required to describe complex events.
What Whorf was really concerned with is the ramifications language
imposes on thought. In his famous example of ‘empty gasoline drums’, it is
the word empty that suggests to the speakers of a language some kind of a
void. Empty means that there is nothing inside, not that they are empty of
gasoline, but contain something else instead, i.e. gasoline vapours.
Revoking the concept of emptiness suggests safety, so the users of the
language habitually think of empty as safe.
Yee (1999) in his interpretation gives another example to illustrate
Whorf’s way of thinking:
Consider the verb ‘to drink’ in English. You can drink coffee, tea,
water, and soup. So it appears that drink has to do with liquids and
it has to do with not chewing whatever is ingested. But we do not drink
all liquids. For example, lighter fluid is ‘fatal if swallowed’, not ‘fatal if
drunk’. This is true for all poisons as well. Also, we ‘swallow’ solid
medicine, but we ‘take’ liquid medicine. So we see that in two special
cases – poisons and liquid medicine, we do not use the word drink. One
may now assume that drink means ‘orally taking some liquid that is
expected to maintain one’s physical well-being’ . . . But even this misses
something, because we can say ‘He drank vodka till he passed out’, and
here we know that drinking tequila does not exactly maintain your
well-being. Another clue to what drinking really refers to is seen in
Socrates’ death. How did Socrates die? People usually say that ‘He died
by drinking hemlock’, not ‘He died by swallowing hemlock’, and here I
think we see more clearly that drinking must also have a volitional
aspect to it – the wanting to ingest something. So drinking is really ‘the
voluntary oral ingestion, without mastication, of liquids, excluding
liquid medicine, with which that one expects to, whether correctly or
not, reach a specific physical state in mind.’ I really don’t think any
dictionary or native speaker would define drink in that way. And that
is exactly Whorf’s point. Because we cannot accurately define such a
common word, we are not even aware of the logical boundaries it
imposes on our formulation of ideas. (Yee 1999)
While the verb drink indeed imposes some ramifications on the way we
formulate our ideas, it still remains to be seen whether such fine specifica-
tions of meaning have cognitive consequences for language users. The
relationship between drinking and swallowing in particular is less clear cut
than the relationship between green and blue, since drinking inevitably
entails swallowing. The question then arises whether the difference
between fatal if swallowed and fatal if drunk is conceptually driven or purely
linguistic and limited to collocability, i.e. the probability of co-occurrence
of certain lexical items.
Cross-linguistic differences in the way events are described go beyond
lexicosemantics. Different languages have different morphosyntactic

386 ANNA EWERT
systems, express tense and aspect differently. In the examples below, three
different Polish verbs correspond to the English verb go:
Ide˛ do domu. – I am going home. (right now)

Chodze˛ do szkoły. – I go to school. (regularly repeated action)
Wychodze˛. – I am going out.
The Polish verbs have two different roots, but all are in the same tense,
while there are two grammatical tenses in English. The question that
cannot readily be answered is whether speakers of English and speakers
of Polish have the same concept of going somewhere, no matter regularly
or not, or perhaps each of the Polish verbs, or roots, corresponds to a
different concept. The question is not trivial, because different conceptual
representations would lead to processing differences, for example in con-
ceptual priming experiments.
In Whorfian-tradition research it has been so far assumed that words
reveal underlying concepts. While this is an easily verifiable assumption in
the domain of colour, there is a possibility that in the area of space and
motion conceptual categories are broader than words. This would be
suggested by the study by Malt et al. (2008), who demonstrate that speakers
of different languages distinguish between walking and running, using
one or more verbs to describe different instances of these two kinds of
human motion. As the authors interpret their results by pointing to some
universals of human perception, this finding might be challenged by
Phelps and Duman (2012), who found that speakers of a Bavarian dialect
of German use the verb laufen to describe instances of both walking and
running.
An alternative proposal, consistent with the relativity hypothesis, is the
semantic cluster hypothesis proposed by Holmes (2012; Holmes and Wolff
2013). In this view, salient conceptual distinctions correspond to elements
of meaning that are shared by multiple words. To identify such semantic
clusters, Holmes and Wolff (2013) asked participants to sort English spatial
prepositions into groups on the basis of similarity. Prominent spatial
relations were then tested in a non-verbal divided visual field task, produ-
cing a categorical perception effect when the stimuli were presented to the
right visual field, similar to the one described above in the domain of
colour (Mo et al. 2011), or in a novel category learning task (Maier et al.
2014). Although only monolingual data in support of the semantic cluster
hypothesis exist so far, speakers of different languages might form differ-
ent semantic clusters, which in turn might lead to differences in non-
linguistic spatial thinking.
A middle-of-the-road perspective on linguistic categorisation, tradition-
ally adopted in motion research, is the one distinguishing different seman-
tic components of event structure such as manner and path (Talmy 1985,
2000). Kersten et al. (2010) used a non-verbal category learning task to test
Slobin’s (2006) manner-salience hypothesis, which assumes that

languages differ with regard to the amount of attention they pay to

manner of motion. In this study the participants were shown animated
cartoons presenting novel events which they had to classify on the basis of
either manner or path. The results showed that speakers of English cate-
gorised on the basis of manner more often than speakers of Spanish, but
there was no difference between the groups in the path category discrimi-
nation tasks. Similar predictions were tested by Czechowska and Ewert
(2011) in two non-verbal tasks, with pictures portraying a human actor
performing different actions. The triad task showed no differences
between monolingual speakers of English and Polish, two typologically
related languages. However, the picture similarity rating task, in which
the participants were not forced to choose between manner and path
alternates, showed that monolingual speakers of Polish pay more
attention to manner than monolingual speakers of English, and Polish–
English bilinguals get desensitised to manner cues even at lower levels of
L2 proficiency. As in the similarity rating task, contrary to the triad task,
both elements of motion, i.e. manner and path, have to be considered at
the same time, it can be argued after Kay and Kempton (1984) that this task
reduces the usefulness of the naming strategy, thus orienting the partici-
pant towards purely perceptual aspects of the visual stimuli. Ewert and
Krzebietke (2015) conducted a follow-up study with the same pool of
participants, collecting motion narratives, to pinpoint the exact differ-
ences in language that might be behind such developments. The results
show, as expected, that monolingual speakers of Polish use manner verbs
more frequently than monolingual speakers of English, the latter using
more neutral verbs, not specified for manner or path. However, if counted
carefully, it turns out that the Polish narratives contain more path
elements overall due to a high incidence of manner verbs with directional
path prefixes (75–80% of all the verbs). The L2 users’ L2 narratives did not
differ from monolingual English narratives with respect to manner
salience, but the L2 users used neutral verbs less frequently, and over-
used path verbs while narrating the events in English. These results
suggest that elements smaller than words can increase conceptual salience
of path under some specific circumstances.
All in all, the research reviewed in this section points to the fact that
semantic categories in the domains of space and motion overlap with
rather than correspond to word boundaries.
18.2 Thinking for speaking
As the discussion so far demonstrates, it is fairly difficult to determine

what constitutes a concept in the domain of motion. Similar operational
concerns led Slobin (1996a) to reconceptualise thought as thinking and
language as speaking. His thinking for speaking is defined as ‘a special

388 ANNA EWERT
form of thought that is mobilized for communication’, more specifically

‘the sort of mental activity that goes on while formulating utterances’,
since ‘We encounter the contents of the mind . . . when they are being
accessed for use’ (Slobin 1996a, p. 76).
Slobin (1996a) notes that there is usually some degree of optionality in
language. Whereas Whorf was concerned with the ramifications that the
structures of a particular language impose on thought, Slobin observes
that speakers usually have a number of options to choose from within the
same language while describing the same situation. Therefore, in his
thinking for speaking approach, he proposes to focus only on those
aspects of experience that have to be obligatorily expressed in a given
language. Thinking for speaking then ‘involves picking those characteris-
tics of objects and events that (a) fit some conceptualisation of the event,
and (b) are readily encodable in the language’ (Slobin 1996a, p. 76).
18.2.1 Whorf and Slobin compared

A brief glance at Whorf’s relativity and Slobin’s thinking for speaking
shows that the two hypotheses deal with different constructs. In tradi-
tional information processing approaches the contents of the mind con-
sists of representations and processes. Representations can be thought of
as some relatively static items of knowledge, while processes are dynamic.
These two are in a complementary relation as processes always act on
representations. Accessing a representation, i.e. accessing some form of
knowledge for use, is a dynamic process. Speech production can be
thought of as a complex operation consisting of a number of more
elementary processes. While Whorf was concerned with mental
representations of reality, Slobin (1996a) steers clear of dealing with the
representation of events, focusing on processing instead. He explicitly
wants to focus on ‘mental processes that occur during the act of formulat-
ing an utterance’ (Slobin 1996a, p. 71) and his object of study is verbalised
events that are constructed online.
Slobin’s (1996a) reluctance to deal with representations was rooted in
his conviction that the perception of motion is universal. He says:
all human beings experience sequences of events that have particular

temporal contours, put objects in locations, and so on. Indeed, animals
do the same . . . Distinctions of aspect, definiteness, voice, and the like, are,
par excellence, distinctions that can only be learned through language, and
have no other use except to be expressed in language. They are not
categories of thought in general, but categories of thinking for speaking.
(Slobin 1996a, p. 91)
Only in his later writings does Slobin (2003, 2006) acknowledge that
repeated attention to some specific conceptual features that need to be
expressed in a particular language might produce different behavioural

outcomes in speakers of different languages. Even in this view, however,

basic cognition and perception of motion does not change. The represen-
tation of motion events stays the same, but certain parts of this represen-
tation become more readily accessible through an attentional bias
produced by preferred language processing strategies. Hence, while both
Whorf and Slobin predict that language may affect speakers’ behaviour,
the locus and the mechanism of this effect is different in both hypotheses.
18.2.2 The processing view

Slobin (2003, 2006) adopts Levelt’s (1989) model of speech production. In
this model, the first stage is conceptualisation of the message that takes
place in the conceptualiser, which produces a pre-verbal message that is
encoded in language at a subsequent processing stage by the formulator.
The nature of the processing that takes place in the conceptualiser has
been elaborated upon by several authors (e.g. Levelt 1999; von Stutterheim
and Nüse 2003; Bylund 2011). In essence, the processes that take place in
the conceptualiser while the preverbal message is being constructed to
describe a real event relate perceptual input from the senses to conceptual
representations in the mind. In brief, the speaker has to make sense of the
observed situation in order to decide which elements of it to encode in
language and how the message is going to be structured and temporally
organised so that appropriate lexical representations can subsequently be
accessed. This kind of processing is envisaged as linear, taking place in a
top-down fashion, where each stage has to be complete for the next one to
begin. What is also important, while Whorf saw a direct link between
conceptual and lexical representations of the kind we can see in psycho-
linguistic models of the bilingual mental lexicon that date back to
Weinreich (1953), Levelt’s (1989, 1999) is a processing model that specifies
the processes that take place before a selected lexical representation is
accessed.
It is assumed after Levelt (1989) that the processing in the conceptualiser
takes place at two levels: macroplanning and microplanning. At the
macroplanning stage the speaker decides what to say. Two processes are
at work here: segmentation and selection (von Stutterheim and Nüse
2003). Segmentation involves matching the observed situation to concep-
tual representations stored in the mind (i.e. the knowledge base) and
breaking it down into states or dynamic events. Selection entails choosing
those elements of the situation that the speaker wants to verbalise. At the
microplanning stage structuring and linearisation of the preverbal
message take place, so that some kind of semantic representation is
activated that can subsequently be encoded in words.
A question that has received some research attention is whether the
processing that goes on in the conceptualiser is affected by the require-
ments of a particular language in which the message is to be encoded.

390 ANNA EWERT
Levelt (1996, 1999) argues that macroplanning is not affected by the spe-
cificities of a particular language, since it is only the preverbal message
that comes out of the conceptualiser that consists of lexical concepts, i.e.
concepts that correspond to actual words in the language. Thus, all the
macroplanning processes that eventually lead to the construction of the
preverbal message operate on knowledge representations that are not
related to language and are probably universal and species-specific. On
the other hand, he equates microplanning with Slobin’s thinking for
speaking, maintaining that it is involved especially when spatial relations
are to be expressed. Von Stutterheim and Nüse (2003) demonstrate that
speakers of English and speakers of German segment the events in their
narrations differently, with English speakers encoding more events, and
they select different components of events to encode, with German speak-
ers encoding the endpoints of motion more often. Hence, the two groups
differ at the macroplanning level of processing. On these grounds von
Stutterheim and Nüse (2003, p. 877) infer that speakers of different lan-
guages apply different principles of information organisation at a pre-
linguistic stage of processing. Principles of information organisation are
language-specific and they function at the pre-linguistic macroplanning
stage as ‘an interface that provides guidelines for the selection and struc-
turing of knowledge that is stored in different formats’.
One weakness of this model is its reliance on explicit verbalisations
from which the processes that led to their production can be inferred.
While linguistic relativity is a hypothesis that can be verified empirically,
thinking for speaking in this formulation is a rationalistic model than
cannot be falsified. It is only rational to assume that something has to be
selected before it is structured and linearised, but we have no means of
verifying if this is the real time sequence of the processes.
18.2.3 Conceptual transfer in second language speakers

A vast portion of research in the thinking for speaking approach relies on
Talmy’s (1985, 2000) distinction between satellite-framed and verb-framed
languages. In satellite-framed languages, such as Germanic or Slavic lan-
guages, verbs convey primarily manner of motion, while path of motion is
encoded in a satellite, i.e. a preposition or a verbal particle. In verb-framed
languages, the main verb in a sentence usually encodes path of motion,
while manner is optional, expressed by adverbial phrases or other verbs.
These relations are frequently referred to as conceptualisation patterns to
emphasise the processes taking place in speech production.
Talmy’s typology turned out to be particularly useful in cross-linguistic
comparisons (e.g. Choi and Bowerman 1991; Slobin 1996b; Hickmann and
Hendriks 2010) as well as in second language acquisition research. This
research typically involves the study of narratives, focusing on how seman-
tic information is distributed in motion event descriptions. Choi and

Bowerman (1991) demonstrate that children become sensitive to lan-

guage-specific patterns of information organisation at the age of 17–20
months. Semantic density of information about motion increases with the
child’s age (Hickmann and Hendricks 2010).
Cross-linguistic comparisons spurred further refinements of the
typology, showing that even within the same typological group languages
differ with respect to how they encode manner and path of motion. For
example, Slavic languages like Russian use manner verbs more frequently
than Germanic languages (Slobin 2006), but English has more manner verb
types (Slobin, Ibarretxe-Antunaño, Kopecka and Majid 2014). Pavlenko and
Volynsky (2015), on the other hand, point out that Talmy’s typology is
insufficient to describe some features of Slavic verbs, such as the obliga-
torily encoded directionality in pairs such as Polish iść – chodzić (‘to walk’),
which refer to unidirectional and non-directed motion respectively.
A theoretical framework to study second language acquisition in the
thinking for speaking perspective is the conceptual transfer hypothesis
(Jarvis and Pavlenko 2008; Bylund and Jarvis 2011). The hypothesis differs
essentially from the predictions stemming from the linguistic relativity
hypothesis by proposing that what is transferred by an L2 user is concep-
tualisation patterns that might reflect either conceptual distinctions or
discursive practices of a particular speech community.
Second language research in the domain of motion shows complex pat-
terns of transfer. The full complexity of these cross-linguistic interactions is
shown by the results obtained by Hohenstein, Eisenberg and Naigles (2006)
with Spanish–English bilinguals. The study participants showed grammati-
cal transfer from L1 to L2, using bare verb utterances in their motion
descriptions, unlike English monolinguals. They also demonstrated reverse
lexical transfer from L2 to L1. Although Hohenstein et al. (2006) state that
the bilinguals’ motion descriptions resemble those of monolinguals in each
language, the bilinguals demonstrated bidirectional transfer, i.e. instances
of L1 influencing the L2 with the L2 influencing the L1 at the same time. A
clear instance of this kind of transfer was the frequency of manner verbs use
in L1 and L2. The bilinguals used fewer manner verbs in English than
English monolinguals under the influence of L1 Spanish, but more manner
verbs in L1 Spanish than Spanish monolinguals.
Second language speakers can also display unique developments, thus
differing from monolingual speakers of both L1 and L2. Brown and
Gullberg (2010) studied the encoding of path in motion event descriptions
of Japanese intermediate learners of English. Japanese being a verb-framed
language, Japanese monolinguals used more path verbs per clause than
English monolinguals, but Japanese intermediate learners of English pro-
vided more path information per clause than any of the monolingual
groups. Another study (Brown and Gullberg 2011) shows that Japanese
learners of English include more goal expressions in their utterances in
both the L1 and the L2 than monolingual speakers of each language.

392 ANNA EWERT
Apart from language, conceptual transfer can also be seen in speech-

accompanying gestures. Brown and Gullberg (2008) studied the encoding
of manner of motion in both speech and gesture by Japanese intermediate
learners of English. While the L2 learners did not differ in the expression of
manner in speech from monolingual speakers of Japanese, they differed
from the Japanese monolinguals by employing manner modulation more
often, that is, they expressed manner of motion in speech, but not in the
speech-accompanying gesture, thus behaving in an L2 appropriate way
while describing motion in the L1.
As the studies conducted by Brown and Gullberg involved intermediate
learners of English, it seems that motion event conceptualisations are very
volatile and susceptible to change as a result of bilingual experience.
A related effect was demonstrated by Daller, Treffers-Daller and Furman
(2011) with Turkish–German bilinguals. They tested two groups of bilin-
guals: one consisted of bilinguals living in Germany; the other consisted of
Turkish–German bilinguals raised in Germany who returned to Turkey.
The bilinguals living in Germany demonstrated German-like conceptuali-
sation patterns in both languages, while the returnees displayed Turkish–
like conceptualisation patterns in both German and Turkish.
When the semantic content of bilingual motion descriptions is studied,
it turns out that the conceptualisation patterns in their L1 and L2 can be
either divergent, i.e. similar to conceptualisation patterns of the mono-
lingual speakers of the respective languages (Hohenstein et al. 2006;
Pavlenko and Volynsky 2015), or convergent, which means that L2 users
display similar conceptualisation patterns in their L1 and L2 and at the
same time they differ from monolingual speakers of each language (Brown
and Gullberg 2008, 2013; Brown 2015).
Another strand of research with second language speakers in the think-
ing for speaking and conceptual transfer framework takes as a starting
point the so-called aspect hypothesis. Languages differ in the way they
express aspectual distinctions. Some languages encode ongoingness, i.e.
the present aspect; some do not make any aspectual distinctions in the
present tense. If ongoingness is encoded in the verb, the focus is on
the action that is taking place. If this aspectual distinction is not made in
the verb, other linguistic means have to be employed to indicate if the
event is bounded. It is assumed that speakers of languages that do not have
grammatical aspect will encode motion endpoints more frequently to indi-
cate the temporal contours of the motion event. Indeed, research shows that
speakers of German, a non-aspect language, mention endpoints of single
events more frequently than speakers of English and Algerian Arabic, both
being aspect languages (von Stutterheim and Nüse 2003).
This line of research also shows an L2 effect on L1. A study conducted
with Spanish–Swedish bilinguals (Bylund and Jarvis 2011), Swedish being
a non-aspect language, shows that the bilinguals mention endpoints in
their L1 Spanish more frequently than Spanish monolinguals, Spanish

being an aspect language. The results of a grammaticality judgement test

show that the frequency of endpoint encoding correlates with the num-
ber of errors made on aspectual distinctions in Spanish, thus providing
evidence for a link between the knowledge of L1 aspectual distinctions
and the frequency of endpoint encoding. Bylund (2009) demonstrates a
link between the age of L2 Swedish acquisition and the frequency of
endpoint encoding in L1 Spanish. The argument here is that aspectual
distinctions are acquired relatively late in L1, so if L2 acquisition starts
before the L1 distinctions are acquired, the influence of the L2 on the L1 is
more pronounced.
Since the thinking for speaking hypothesis emphasises online proces-
sing in formulating an utterance, endpoint encoding has also been studied
with the use of the eye-tracking technique. Eye tracking records fixations
with fine temporal resolution. Therefore it is a useful technique for exam-
ining attention to different elements of a situation at a speech planning
stage, particularly useful for researching the thinking for speaking
hypothesis. Von Stutterheim et al. (2012) propose a seeing for speaking
hypothesis, assuming that speakers will attend to visual features of a
frequently and systematically used grammaticised concept, but not to
features of a lexically encoded concept. Research with speakers of seven
languages shows that speakers of non-aspect languages look towards
endpoints of ongoing motion events more often than speakers of aspect
languages. Subsequent eye tracking research with French highly
proficient users of L2 English (Flecken, Carroll, Weimar and von
Stutterheim 2015) shows that the L2 users learn to attend to manner
information in the way German native speakers do, but their L1 path
information processing preferences and spatial frames as revealed by
gaze allocation patterns are deeply entrenched, so that even when speak-
ing their L2 German, they attend to path information in the visually
presented scenes in a way similar to French monolinguals.
18.3 Embodied cognition
The third theoretical perspective in motion research is the embodied (or

grounded) cognition hypothesis (Glenberg and Kaschak 2002; Wilson
2002; Barsalou 2008). By rejecting the view that language is a system of
amodal, arbitrary signs, this hypothesis proposes that linguistic meanings
are embodied, which means that they are processually linked to the
sensory and biomechanical experiences of an individual. In the case of
motion and action language in general, this literally means that linguistic
representations of movement are connected to sensory-motor representa-
tions of physical action. Neurophysiological evidence confirms that action
verbs indeed activate the motor cortices of the brain (Pulvermüller, Härle
and Hummel 2001; Tettamanti et al. 2005). Behavioural research shows

394 ANNA EWERT
that action language facilitates physical responses if the linguistically

encoded and bodily movement directions are congruent, and interferes
with physical response if both directions are incongruent (Glenberg and
Kaschak 2002). These findings are frequently explained as mental simula-
tion, i.e. the mind simulates, or re-enacts, the action encoded in language.
Numerous theoretical accounts of embodiment have been proposed
(Wilson 2002; Barsalou 2008). For example, Barsalou (1999) postulates
that action is mentally represented as perceptual symbol systems (PSS),
thus proposing some form of symbolic representation. On the other hand,
Simmons et al. (2008) propose the language and situated simulation theory
(LASS). In this view, linguistic and conceptual processing always take
place simultaneously, but both systems are differentially involved depend-
ing on the task at hand, which indicates that conceptual processing relies
on multiple representations.
Neurological research provides further evidence that different areas of
the brain are involved in spatial thought and language. Wu, Morganti and
Chatterjee (2008) provide evidence for parallel neural organisation of
spatial information and language, with different regions of the brain
responsible for processing path and manner information. Desai et al.
(2013) demostrate that motion sentences involve the motor cortices of
the brain in different ways depending on the context in which the motion
is embedded. Sentences describing physical motion activate primary
motor cortices as well as areas implicated in planning physical movement.
Figurative motion sentences containing the same verb do not activate
primary motor cortices but engage areas involved in higher-order motor
processes.
Motor activations are further modulated by grammar. Zwaan, Taylor and
de Boer (2010) found that motor systems are involved in the processing of
descriptions of present and past actions, but not in the processing of action
intentions. Bergen and Wheeler (2010) demonstrated that the activation of
motor systems is modulated by grammatical aspect.
The focus on how the brain grounds sensory and motor activations
creates an impression that these patterns of neural activations correspond
somehow to conceptual representations. While this indeed might be the
case, we are still far from understanding the exact nature of conceptual
representations. It is generally agreed that concepts are distributed repre-
sentations whose features are activated depending on contextual con-
straints, but some models assume that there might be a region in the
brain where this distributed sensory-motor information is integrated
(Kiefer and Pulvermüller 2012; Meteyard et al. 2012).
The embodied cognition perspective, proposing an entirely different
perspective on the nature of human cognition, opens a number of
vistas for both cross-linguistic and bilingualism research. First of all,
Bergen et al. (2010, Experiment 4) asked the question whether non-
native speakers activate motor representations in their L2. The

participants were shown pictures presenting different actions and

verbs that described either the action portrayed in the picture or a
different action. The different actions could be performed with either
the same effector, i.e. the same part of the body, or a different effector.
The participants’ task was to judge if the verbs provided a good descrip-
tion of the corresponding picture. When the ‘different’ verbs involved
the same effector, the participants’ responses were slower, both native
and non-native speakers’. Also de Grauwe et al. (2014) show that
advanced L2 speakers’ motion representations are embodied.
A different question has been asked by Tomczak and Ewert (2015). Since
it is known that context modulates the involvement of motor systems, the
question was if speakers of different languages would show the same
effects in a conceptual priming experiment. Two groups of monolinguals
and a group of Polish L2 users of English were asked to judge the mean-
ingfulness of real and fictive motion sentences preceded with an action
verb or a noun prime. In the experiment, English monolingual partici-
pants showed the expected priming effect, but not the Polish monolin-
guals, and the L2 users demonstrated a pattern of cross-linguistic influence
in their conceptual processing. While it is not exactly clear why the Polish
monolinguals did not show any effect of priming, one quite obvious
possibility being that perhaps the task was not sensitive enough to detect
embodiment effects in speakers of Polish, this result indicates that the way
different languages structure information might affect availability of some
part of conceptual representation. This at least is the explanation sug-
gested by the findings on tense (Zwaan, Taylor and de Boer 2010) and
aspect (Bergen and Wheeler 2010) processing cited above. It would be
interesting to see how cross-linguistic differences in word order or deriva-
tional morphology affect availability of sensory-motor information in
linguistic processing and how they affect bilingual cognition.
18.4 Conclusion
Space and motion research has been conducted within three different
theoretical frameworks. The thinking for speaking approach has produced
the largest amount of bilingual research by providing a convenient typol-
ogy to be used in this type of comparisons, but all three approaches are
vibrant and open up interesting lines for future investigation by focusing
on different constructs.
Although there are certain similarities between these different
approaches and research relying on different theoretical assumptions
frequently makes use of the same typology of motion events proposed by
Talmy, these three orientations make different assumptions about the
locus and mechanism of the influence of language on space and motion
cognition, or the nature of spatial thought.

396 ANNA EWERT
The linguistic relativity approach assumes some separation between

linguistic and conceptual representations and a causal effect of language
on conceptual categories. The relationship between linguistic and concep-
tual categories in the thinking for speaking approach is obscured by its
focus on online processing and conceptualisations, which are a product of
these online processes. The thinking for speaking approach focuses on
online speech planning processes and implicates an attentional mechan-
ism of the influence of these processes on motion cognition. In the
embodiment perspective linguistic and sensory-motor representations
are integrated and inseparable.
These different theoretical foci lead to different research questions that
make use of different research techniques. Recent research in the linguis-
tic relativity framework focuses on categorical perception using psycho-
linguistic and electrophysiological measures, trying to answer the
questions relating to which categorical distinctions are most salient and
what is the nature of conceptual categories. The most promising line of
research in the thinking for speaking perspective seems to be an eye-
tracking study of cross-linguistic differences in speech planning processes.
In the embodiment perspective, imaging techniques are used to elucidate
how the brain makes sense of spatial and sensory-motor information.
Behavioural experiments and electrophysiological studies throw light
onto how this information is processed.
Second language research in the domain of space and motion has so far
relied mostly on the thinking for speaking approach, with a limited num-
ber of studies explicitly adhering to the relativity assumption and only a
few single studies in the embodiment tradition. We know by now that
second language users do embody motion information. We also know that
bilingual cognition is extremely malleable and susceptible to different
influences. It would be interesting to see how cross-linguistic differences
in the way motion is encoded in language modulate the way L2 users
ground sensory-motor and spatial information.
References

9, 89–96.
Athanasopoulos, P. and Bylund, E. 2013. ‘Does grammatical aspect affect
motion event cognition? A cross-linguistic comparison of English and
Swedish speakers’, Cognitive Science 37, 286–309.
Barsalou, L. W. 1999. ‘Perceptual symbol systems’, Behavioural and Brain
Sciences 22, 577–660.
Barsalou, L. W. 2008. ‘Grounded cognition’, Annual Review of Psychology 59,
617–645.

Bergen, B., Lau, T.-T. C., Narayan, S., Stojanovic, D. and Wheeler, K. 2010.
‘Body part representations in verbal semantics’, Memory and Cognition 38,
969–981.
Bergen, B. and Wheeler, K. 2010. ‘Grammatical aspect and mental simula-
tion’, Brain and Language 112, 150–158.
Berlin, B. and Kay, P. 1969. Basic Color Terms. Berkeley, CA: University of
California Press.
Boroditsky, L., Ham, W. and Ramscar, M. 2002. ‘What is universal in event
perception? Comparing English & Indonesian speakers’, in W. D. Gray
and C. D. Schunn (eds), Proceedings of the 24th Annual Conference of the
Cognitive Science Society, pp. 136–141. Lawrence Erlbaum.
Brown, A. 2008. ‘Gesture viewpoint in Japanese and English: cross-linguistic
interactions between two languages in one speaker’, Gesture 8, 256–276.
Brown, A. 2015. ‘Universal development and L1–L2 convergence in bilin-
gual construal of manner in speech and gesture in Mandarin, Japanese,
and English’, Modern Language Journal 99 (S1), 66–82.
Brown, A. and Gullberg, M. 2008. ‘Bidirectional crosslinguistic influence in
L1–L2 encoding of manner in speech and gesture: a study of Japanese
speakers of English’, Studies in Second Language Acquisition 30, 225–251.
Brown, A. and Gullberg, M. 2010. ‘Changes in encoding of PATH of motion
in a first language during acquisition of a second language’, Cognitive
in event conceptualization? Expressions of Path among Japanese lear-
ners of English’, Bilingualism: Language and Cognition 14, 79–94.
Brown, A. and Gullberg, M. 2013. ‘L1–L2 convergence in clausal packaging
in Japanese and English’, Bilingualism: Language and Cognition 16, 477–494.
Brown, R. W. and Lenneberg, E. H. 1954. ‘A study in language and cogni-
tion’, Journal of Abnormal and Social Psychology 49, 454–462.
Bylund, E. 2009. ‘Effects of age of L2 acquisition on L1 event conceptualiza-
tion patterns’, Bilingualism: Language and Cognition 12, 305–322.
Bylund, E. 2011. ‘Language-specific patterns in event conceptualization:
insights from bilingualism’, in A. Pavlenko (ed.), Thinking and Speaking in
Two Languages, pp. 108–142. Bristol: Multilingual Matters.
Carroll, J. B. (ed.) 1956. Language, Thought, and Reality: Selected Writings of
Benjamin Lee Whorf. Cambridge, MA: MIT Press.
Casasanto, D. 2008. ‘Who’s afraid of the Big Bad Whorf? Crosslinguistic
differences in temporal language and thought’, Language Learning 58
(S1), 63–79.
Choi, S. and Bowerman, M. 1991. ‘Learning to express motion events in
English and Korean: the influence of language-specific lexicalization
patterns’, Cognition 41, 83–121.

398 ANNA EWERT
Cook, R. S., Kay, P. and Regier, T. 2005. ‘The World Color Survey database:
history and use’, in H. Cohen and C. Lefebvre (eds), Handbook of
Categorization in the Cognitive Sciences, pp. 223–242. Elsevier.
557–591.
Cook, V. J. 2003. ‘The changing L1 in the L2 user’s mind’, in V. J. Cook (ed.),
Effects of the Second Language on the First, pp. 1–18. Clevedon: Multilingual
Matters.
Japanese L2 users of English’, International Journal of Bilingualism 10, 137–152.
Costa, A., Strijkers, K., Martin, C. and Thierry, G. 2009. ‘The time course of
word retrieval revealed by event-related brain potentials during overt
speech’, Proceedings of the National Academy of Sciences 50, 21442–21446.
Czechowska N. and Ewert, A. 2011. ‘Perception of motion by Polish–
English bilinguals’, in V. J. Cook and B. Bassetti (eds), Language and
Bilingual Cognition, pp. 287–314. New York and Hove: Psychology Press.
Daller, M. H., Treffers-Daller, J. and Furman, R. 2011. ‘Transfer of concep-
tualization patterns in bilinguals: the construal of motion events in
Turkish and German’, Bilingualism: Language and Cognition 14, 95–119.
De Grauwe, S., Willems, R. M., Rueschemeyer, S.-A., Lemhöfer, K. and
Schriefers, H. 2014. ‘Embodied language in first- and second-language
speakers: neural correlates of processing motor verbs’, Neuropsychologia
56, 334–349.
Desai, R. H., Conant, L. L., Binder, J. R., Park, H. and Seidenberg, M. S. 2013.
‘A piece of the action: modulation of sensory-motor regions by action
idioms and metaphors’, Neuroimage 83, 862–869.
Ewert, A. and Krzebietke, W. 2015. ‘Manner and path of motion in descrip-
tions of motion trajectories by Polish L2 users of English’, in L. Roberts,
K. McManus, N. Vanek and D. Trenkic (eds), EUROSLA Yearbook, vol. 15,
Flecken, M., Carroll, M., Weimar, K. and von Stutterheim, C. 2015. ‘Driving
along the road or heading for the village? Conceptual differences under-
lying motion event encoding in French, German, and French–German L2
users’, Modern Language Journal 99 (S1), 100–122.
Flecken, M., von Stutterheim, C. and Carroll, M. 2014. ‘Grammatical aspect
influences motion event perception: findings from a cross-linguistic
nonverbal recognition task’, Language and Cognition 6, 45–78.
Gennari, S. P., Sloman, S. A., Malt, B. C. and Fitch, W. T. 2002. ‘Motion
events in language and cognition’, Cognition 83, 49–79.
Gilbert, A. L., Regier, T., Kay, P. and Ivry, R. B. 2008. ‘Support for lateraliza-
tion of the Whorf effect beyond the realm of color discrimination’, Brain
and Language 105, 91–98.

Glenberg, A. M. and Kaschak, M. P. 2002. ‘Grounding language in action’,

Psychonomic Bulletin and Review 9, 558–565.
Hedden, T., Ketay, S., Aron, A., Markus, H. R. and Gabrieli, J. D. E. 2008.
‘Cultural influences on neural substrates of attentional control’,
Psychological Science 19, 1–17.
Hickmann, M. and Hendricks, H. 2010. ‘Typological constraints on the
acquisition of spatial language in French and English’, Cognitive
Hohenstein, J., Eisenberg, A. and Naigles, L. 2006. ‘Is he floating across or
crossing afloat? Cross-influence of L1 and L2 in Spanish–English bilin-
gual adults’, Bilingualism: Language and Cognition 9, 249–261.
Holmes, K. J. 2012. Language as a Window into the Mind: The Case of Space.
Doctoral dissertation, Emory University.
Holmes, K. J. and Wolff, P. 2013. ‘When is language a window into the
mind? Looking beyond words to infer conceptual categories’, in M.
Knauff, M. Pauen, N. Sebanz and I. Wachsmuth (eds), Proceedings of the
35th Annual Conference of the Cognitive Science Society, pp. 597–602. Cognitive
Science Society.
Cognition. New York: Taylor and Francis.
Kay, P., to appear. ‘Universality of color categorization’, in A. J. Elliot, M. D.
Fairchild and A. Franklin (eds), Handbook of Color Psychology. Cambridge
University Press.
Kay, P. and Kempton, W. 1984. ‘What is the Sapir–Whorf hypothesis?’,
American Anthropologist 86, 65–79.
Kersten, A. W., Meissner, C. A., Lechuga, J., Schwartz, B. L., Albrechtsen, J. S.
and Iglesias, A. 2010. ‘English speakers attend more strongly than
Spanish speakers to manner of motion when classifying novel objects
and events’, Journal of Experimental Psychology: General 139, 638–653.
Kiefer, M. and Pulvermüller, F. 2012. ‘Conceptual representations in mind
and brain: theoretical developments, current evidence and future direc-
tions’, Cortex 48, 805–825.
Levelt, W. J. M. 1989. Speaking: From Intention to Articulation. Cambridge, MA:
MIT Press.
Levelt, W. J. M. 1996. ‘Perspective taking and ellipsis in spatial descrip-
tions’, in P. Bloom, M. Peterson, L. Nadel and M. Garrett (eds), Language
and Space, pp. 77–107. Cambridge MA: MIT Press.
Levelt, W. J. M. 1999. ‘Producing spoken language: a blueprint for the
speaker’, in C. Brown and P. Hagoort (eds), The Neurocognition of
Language, pp. 83–120. Oxford University Press.
Levinson, S. C. 1996. ‘Frames of reference and Molyneux’s question: cross-
linguistic evidence’, in P. Bloom, M. Peterson, L. Nadel and M. Garrett
(eds), Language and Space, pp. 109–169. Cambridge, MA: MIT Press.
Levinson, S. C. 2003. Space in Language and Cognition: Explorations in Linguistic
Diversity. Cambridge University Press.

400 ANNA EWERT
Lucy, J. A. 1997. ‘Linguistic relativity’, Annual Review of Anthropology 26,

291–312.
Maier, M., Glage, P., Hohlfeld, A. and Abdel Rahman, R. 2014. ‘Does the
semantic content of verbal categories influence categorical perception?
An ERP study’, Brain and Cognition 91, 1–10.
Malt, B. C., Gennari, S., Imai, M., Ameel, E., Tsuda, N. and Majid, A. 2008.
‘Talking about walking: biomechanics and the language of locomotion’,
Psychological Science 19, 232–240.
Meteyard, L., Rodriguez Cuadrado, S., Bahrami, B. and Vigliocco, G. 2012.
‘Coming of age: a review of embodiment and the neuroscience of seman-
tics’, Cortex 48, 788–804.
Mo, L., Xu, G., Kay, P. and Tan, L.-H. 2011. ‘Electrophysiological evidence for
the left-lateralized effect of language on preattentive categorical perception
of color’, Proceedings of the National Academy of Sciences 108 (34), 14026–14030.
Papafragou, A., Hulbert, J. and Trueswell, J. 2008. ‘Does language guide
event perception? Evidence from eye movements’, Cognition 108,
155–184.
Papafragou, A., Massey, C. and Gleitman, L. 2002. ‘Shake, rattle, ’n’ roll: the
representation of motion in language and cognition’, Cognition 84,
189–219.
Pavlenko, A. 2005. ‘Bilingualism and thought’, in J. F. Kroll and
A. M. B. de Groot (eds), Handbook of Bilingualism: Psycholinguistic
Approaches, pp. 433–453. Oxford University Press.
Pavlenko, A. 2011. ‘Introduction: bilingualism and thought in the 20th
century’, in A. Pavlenko (ed.), Thinking and Speaking in Two Languages,
Pavlenko, A. and Volynsky, M. 2015. ‘Motion encoding in Russian and
English: moving beyond Talmy’s typology’, Modern Language Journal 99
(S1), 32–48.
Phelps, K. S. and Duman, S. 2012. ‘Manipulating manner: semantic repre-
sentations of human locomotion verbs in English and German’, in N.
Miyake, D. Peebles and R. P. Cooper (eds), Proceedings of the 34th Annual
Conference of the Cognitive Science Society, pp. 857–863. Cognitive Science
Society.
Pulvermüller, F., Härle, M. and Hummel, F. 2001. ‘Walking or talking?
Behavioral and neurophysiological correlates of action verb processing’,
Brain and Language 78, 143–168.
Roberson, D. and Davidoff, J. 2000. ‘The categorical perception of colors
and facial expressions: the effect of verbal interference’, Memory and
Segall, M. H., Campbell, D. T. and Herskovits, M. J. 1966. The Influence of
Culture on Visual Perception. Bobbs-Merrill.
Simmons, W. K., Hamann, S. B., Harenski, C. N., Hu, X. P. and Barsalou,
L. W. 2008. ‘fMRI evidence for word association and situated simulation
in conceptual processing’, Journal of Physiology: Paris 102, 106–119.

Slobin, D. I. 1996a. ‘From “thought and language” to “thinking for speak-

ing” ’, in J. J. Gumperz and S. C. Levinson (eds), Rethinking Linguistic
Slobin, D. I. 1996b. ‘Two ways to travel: verbs of motion in English and
Spanish’, in M. Shibatani and S. A. Thompson (eds), Grammatical
Constructions: Their Form and Meaning, pp. 195–219. Oxford: Clarendon Press.
Slobin, D. I. 2003. ‘Language and thought online: cognitive consequences
of linguistic relativity’, in D. Gentner and S. Goldin-Meadow
(eds), Language in Mind: Advances in the Study of Language and Thought,
Slobin, D. I. 2006. ‘What makes manner of motion salient? Explorations in
linguistic typology, discourse, and cognition’, in M. Hickmann and S.
Robert (eds), Space in Languages: Linguistic Systems and Cognitive Categories,
pp. 59–81. Amsterdam and Philadelphia: John Benjamins.
Slobin, D. I., Ibarretxe-Antuñano, I., Kopecka, A. and Majid, A. 2014.
‘Manners of human gait: a crosslinguistic event-naming study’,
Cognitive Linguistics 25, 701–741.
Talmy, L. 1985. ‘Lexicalization patterns: semantic structure in lexical
forms’, in T. Shopen (ed.), Language Typology and Syntactic Description, vol.
3, pp. 57–149. Cambridge University Press.
Talmy, L. 2000. Toward a Cognitive Semantics, vol. 2: Typology and Process in
Concept Structuring. Cambridge, MA: MIT Press.
Tettamanti, M., Buccino, G., Saccuman, M. C., Gallese, V., Danna, M., Scifo,
P., Fazio, F., Rizzolatti, G., Cappa, S. F. and Perani, D. 2005. ‘Listening to
action-related sentences activates fronto-parietal motor circuits’, Journal
of Cognitive Neuroscience 17, 273–281.
Thierry, G., Athanasopoulos, P., Wiggett, A., Dering, B. and Kuipers, J.-R.
2009. ‘Unconscious effects of language-specific terminology on pre-
attentive color perception’, Proceedings of the National Academy of Sciences
106, 4567–4570.
Tomczak, E. and Ewert, A. 2015. ‘Real and fictive motion processing in
Polish L2 users of English and monolinguals: evidence for different
conceptual representations’, Modern Language Journal 99 (S1), 49–65.
Trueswell, J. C. and Papafragou, A. 2010. ‘Perceiving and remembering
events cross-linguistically: evidence from dual-task paradigms’, Journal
of Memory and Language 63, 64–82.
Von Stutterheim, C., Andermann, M., Carroll, M., Flecken, M. and
Schmiedtovà, B. 2012. ‘How grammaticized concepts shape event con-
ceptualization in language production: insights from linguistic analysis,
eye tracking data, and memory performance’, Linguistics 50, 833–867.
Von Stutterheim, C. and Nüse, R. 2003. ‘Processes of conceptualization
in language production: language-specific perspectives and event
construal’, Linguistics 41, 851–881.
Mouton.

402 ANNA EWERT
Whorf, B. L. 1940a/1956. ‘Linguistics as an exact science’, reprinted in J. B.

Carroll (ed.) 1956, pp. 220–232.
Whorf, B. L. 1940b/1956. ‘Science and linguistics’, reprinted in J. B. Carroll
(ed.) 1956, pp. 207–219.
Whorf, B. L. 1941/1956. ‘The relation of habitual thought and behavior to
language’, reprinted in J. B. Carroll (ed.) 1956, pp. 134–159.
Wilson, M. 2002. ‘Six views of embodied cognition’, Psychonomic Bulletin and
Review 9, 625–636.
Wu, D. H., Morganti, A. and Chatterjee, A. 2008. ‘Neural substrates of
processing path and manner information of a moving event’,
Yee, N. 1999. What Whorf really said. Manuscript. www.nickyee.com/pon
der/whorf.html
Zwaan, R. A., Taylor, L. J. and de Boer, M. 2010. ‘Motor resonance as a
function of narrative time: further tests of the linguistic focus hypoth-
esis’, Brain and Language 112, 143–149.

19
Multi-competence
and personality
Jean-Marc Dewaele
19.1 Introduction
In a recent contribution on the origin of the term ‘multi-competence’,

Cook (2012) reminds the reader of the Chomskyan influence in the
original definition: ‘the compound state of a mind with two grammars’
(Cook 1991), where ‘grammar’ was used ‘in the sense of the total knowl-
edge of language in the mind (the I-language)’ (Cook 2012, p. 3768).
Cook’s current working definition is ‘the overall system of a mind or a
community that uses more than one language’ (Chapter 1, this volume,
p. 3). Cook’s innovation in the world of applied linguistics was that the
second language (L2) user was considered as a whole person with knowl-
edge of two languages – and the level of proficiency in the L2 was largely
irrelevant. In this respect he further explored the wholistic view of
bilingualism that Grosjean (1989) had put forward, namely that since
bilinguals are not the sum of two monolinguals, they cannot be studied
like any monolingual. Indeed, bilinguals have ‘a unique and specific
linguistic configuration’ (p. 3).
Cook’s wholistic view of bilingualism and his definition of multi-
competence are cognitive. Indeed, Cook (2012) explains that it
is ‘neither particularly a psychological concept, as some have
claimed . . ., nor particularly sociological’ (p. 3768). Instead it focuses
on the mind: ‘Multi-competence therefore involves the whole mind of
the speaker, not simply their first language (L1) or their second’
(p. 3768). However, Cook has never restricted multi-competence to the
linguistic realm, agreeing that the acquisition of an L2 can have non-
linguistic consequences: ‘Acquiring another language alters the L2
user’s mind in ways that go beyond the actual knowledge of language
itself’ (Cook 2002, p. 7). Cook insisted that multi-competence is not
‘a model nor a theory so much as an overall perspective or framework’
(2002, p. 1). While this is a strength, it is also a limitation since it is

404 JEAN-MARC DEWAELE
hard to falsify a perspective or framework, and impossible to quantify

‘multi-competence’. At best the concept can be invoked to explain the
effect of a certain degree of multilingualism on dependent variables.
Few researchers would currently reject Cook’s bilingual ‘wholistic’
interpretation of bilingualism. Indeed the monolingual ‘fractional’
interpretation of bilingualism, in Grosjean’s (1989, 2008) terms has
been largely relegated to the past.
It is slightly surprising that the concept of multi-competence, which
has become really popular in applied linguistics, has remained untouched
by psychologists1 (as argued in Chapter 4 of this volume). The lack of
interest among psychologists could be linked to their general lack of
interest in applied linguistics and multilingualism.2
Also, the fact that Cook never presented a clearly operationalisable
definition of multi-competence has probably also hindered its uptake
among psychologists. Of course, one could argue that a paradigm shift is
not operationalisable or indeed testable in the same way as a model; it
simply means that everything has to be reconsidered from a different
perspective.
We have argued that the growth of individual multilingualism, and the
resulting multi-competence, affects not just an individual’s cognition but
also that individual’s personality (Dewaele and Li Wei 2012, 2013a). This is
again a perspective that has been absent among personality psychologists,
for whom personality is more the result of the workings of nature than of
nurture (McCrae et al. 2000; Pervin and Cervone 2010). Although person-
ality psychologists agree that long-term social and environmental factors
can contribute in shaping an individual’s personality traits, there is
surprisingly little research on this. One possible explanation for this lack
is the widely shared assumption among personality psychologists that
their participants are monolingual and monocultural, and that the pre-
sence of other languages can be safely ignored. Pavlenko notes that this
‘monolingual’ view is not just widely shared among psychologists, but is
also prevalent among linguists and anthropologists (2005, p. 3).
Applied linguists feel that individuals’ multilingualism and multicultur-
alism can shape their personality. Regular exposure to different languages,
values, rituals and practices makes the multilingual a born traveller, and
as Mark Twain pointed out in The Innocents Abroad: ‘Travel is fatal to
prejudice, bigotry, and narrow-mindedness . . .’ (1869, p. 650).
Section 19.2 will consider the tricky issues of definition and operationa-
lisation of multilingualism. Section 19.3 will present research that
combined psychology and multilingualism research. A brief section will
be devoted to the research of cross-cultural psychologists who used
personality traits to predict behaviour of multilinguals. After that the
focus will shift to the studies where psychological variables were the
dependent variables, with multilingualism as the independent variables
(Sections 19.3.1–19.3.3). Some of these dependent variables are proper

Multi-competence and personality 405
personality traits, others are psychological concepts (divergent thinking,

creativity, language aptitude) or affective states (communicative and
foreign language anxiety, foreign language enjoyment).
19.2 Measuring and operationalising multilingualism
Wang (2008) points to the tricky terminological problems concerning the

measurement and the labelling of multilinguals’ languages. Multilinguals’
languages are usually numbered chronologically (L1, L2, L3 . . .), according
to the first exposure to the languages (Hammarberg 2009). This is fine
when individuals grew up as monolinguals and acquired foreign languages
consecutively. It becomes more complicated for children growing up with
two, three, or four languages simultaneously. How to define the first
‘foreign’ language learnt at school for a child with three ‘first’ languages,
L1a, L1b, L1c – is it their ‘L2’, or ‘L4’? Another problem is that of inter-
mittent or alternating acquisition (Hammarberg 2009, p. 4) when a first
relatively short period of exposure might have been interrupted for a
while, with renewed exposure much later. Hammarberg (2009) also raises
the question of limited knowledge of a language and the point at which
multilinguals can include it in their repertoire. Opinions typically differ on
the amount of proficiency one should have before claiming to master a
language. It is equally unclear whether dead languages should be included
in the total language count. The knowledge of such languages (Latin,
ancient Greek, Sanskrit) could have implications for the acquisition of
modern languages – I personally remember the delight of recognising
Latin conjugations when studying Spanish.
The labels L1, L2, L3, L4, L5 and the total language counts differ accord-
ing to researchers’ research orientation. The ‘L1’ is typically defined as the
language(s) that was (were) established up to a certain level before the age
of three. For some researchers the label ‘L2’ refers to any language
acquired after that. It then becomes an umbrella term for all ‘foreign’
languages (Mitchell and Myles 1998, p. 2). Cook (2002) also used the word
‘second’3 as an umbrella term while insisting that the proficiency of L2
users in their different languages may vary widely:
Some of them use the second language as skilfully as a monolingual native
speaker, like Nabokov writing whole novels in a second language; some
of them can barely ask for a coffee in a restaurant. (Cook 2002, p. 3)
Cook felt he did not need a finer-grained distinction between the lan-
guages acquired after the L1(s), and thus labels them all L2s. As he would
typically not establish a total language count for a multilingual, this
mattered little.
Researchers who have focused on the acquisition of a new foreign
language after the second language have argued that this acquisition

may be qualitatively different from the acquisition of the L2 and that a

different term such as ‘L3’ was in order (Hammarberg 2009). Many
researchers in trilingualism thus use the term ‘third’ as an umbrella
term referring to ‘third or additional languages’ (de Angelis 2007).
My own interest in adult individuals who have learnt second, third,
fourth, fifth and more languages at various times during their life meant
that I have avoided umbrella terms. However, establishing a total number
of languages for an individual is fraught with problems. Even widely used
terms such as ‘bilingual’ might be perceived differently by laypeople and
linguists. We were struck by the range of interpretations of the concept of
bilingualism when promoting our Bilingualism and Emotions online web
questionnaire (Dewaele and Pavlenko 2001–2003). Many non-linguist bi-
and multilingual friends and colleagues whom we invited to fill out the
questionnaire declined, saying they did not consider themselves to be
bilinguals, or multilingual enough, even though we felt they qualified. It
could of course have been an excuse to escape a relatively time-consuming
exercise in metapragmatic awareness, but it could also have been the
illustration of the gap between linguists’ current understanding of the
concept of bi- and multilingualism and the laypersons’ view (including
many linguists and psychologists).
Dewaele, Housen and Wei (2003) have argued, like Grosjean and Cook,
in favour of a broad definition of bilingualism that includes: ‘not only the
“perfect” bilingual (who probably does not exist) or the “balanced” ambi-
lingual (who is probably rare) but also various “imperfect” and “unstable”
forms of bilingualism, in which one language takes over from the other(s)
on at least some occasions and for some instances of language use’
(2003, p. 1).
Thompson (2013) has argued that it is very difficult to operationalise the
concepts of bilingualism and multilingualism because of the interplay
of inter-related factors, including proficiency in the various languages,
language choices in everyday life, the relationship between the L1 or
other languages studied and the context of language acquisition.
Researchers also face a practical problem when using stringent criteria:
‘the pool of potential participants inevitably decreases, especially when
dealing with a classroom setting for recruiting purposes’ (p. 697).
Any calculation of ‘total number of languages known’ should therefore
include partly mastered languages. However, such a label might be general
to the point of being less useful for research. In order to obtain a little bit
more granularity, Dewaele and Stavans (2014) developed measures includ-
ing information on proficiency and frequency of use of the various
languages. A first multilingualism index based on language knowledge,
or a ‘total proficiency score’, which is the sum of self-perceived compe-
tence scores (in answer to the question: ‘How proficient are you in your
L1/L2/L3/L4/L5/L6?’) collected on 5-point Likert scales (ranging from 1, mini-
mal proficiency, to 5, maximal proficiency) in up to six languages. Such a

measure is potentially useful to distinguish self-professed pentalinguals

with minimal competence in three languages from a trilingual with
maximal proficiency in three languages. The latter might know fewer
languages, but knowing them to higher level undoubtedly make the indi-
vidual more strongly multilingual. In other words, rather than sticking to
the imprecise labels ‘bilingual, trilingual, quadrilingual etc.’, considering
every language as a discrete entity, despite the fact that competence can be
near zero, we consider the multilingual user’s accumulated language
knowledge across languages. The total proficiency score is the sum of the
proficiency scores on 5-point Likert scales for oral proficiency (maximum
score 5) and written proficiency (maximum score 5) for up to six languages
(including two L1s) (maximal possible score 10 × 6 = 60).
The same principle can be applied for the use of various languages.
Multilinguals who rarely use their foreign languages can be distinguished
from those who use them more frequently. This is a measure of general
intercultural communicative activity. The total language use score is the
sum of frequency of use scores on 5-point Likert scales for up to six
languages (maximal possible score 5 × 6 = 30). The measures were used
in Dewaele and Li Wei (2013b, 2014).
19.3 Personality and multilingualism
Personality traits ‘refer to consistent patterns in the way individuals

behave, feel and think’ (Pervin and Cervone 2010, p. 228). They thus
‘summarise a person’s typical behaviour’ (2010, p. 229). There seems to
be consensus among psychologists about the taxonomy of personality
traits: five broad, bipolar dimensions, the so-called Big Five (openness,
conscientiousness, extraversion, agreeableness and neuroticism) (2010,
p. 228). The Big Five are situated at the summit of the hierarchy; there
are a large number of narrower facets, ‘lower-order’ personality traits, that
are often correlated with the Big Five traits but also explain unique var-
iance. Trait Emotional Intelligence, for example, is formally defined as a
constellation of emotional self-perceptions located at the lower levels of
personality hierarchies (Petrides, Pita and Kokkinaki 2007). It is positively
linked to Extraversion and negatively to Neuroticism but has incremental
validity over the Big Five dimensions in predicting criteria such as depres-
sion, life satisfaction, coping styles and the recognition of emotional
expressions (Davey 2005).
Kihlstrom (2013) explains that personality psychologists measure
individuals’ scores on personality dimensions using questionnaires and
then correlate these with some criterion behaviour in some specific situa-
tion. They generally construe the effects of the environment as ‘noise’
(Kihlstrom 2013). In other words, personality psychologists are more inter-
ested in the physiological sources of personality and much less so in social

factors (McCrae et al. 2000). However, social psychologists, by contrast,

have construed behaviour as a function of differences in the physical and
(especially) social environment (Kihlstrom 2013). A compromise position
emerged with Interactionists, for whom personal and environmental
determinants of behaviour interacted with each other in a variety of
ways (Magnusson and Endler 1977). This paradigm is perfectly suited for
applied linguists who are keen to find out whether multi-competence
reshapes personality profiles.
19.3.1 The effect of personality on adjustment and language use

Cross-cultural psychologists4 have focused on the personality traits
associated with positive outcomes of immigration and psychological adap-
tation in the host country (Kim 2001; Chen, Benet-Martı́nez and Harris
Bond 2008). While these researchers are more interested in the predictive
value of personality, they note that this could be a chicken and egg
situation. Indeed, those who decide to move abroad or go on exchange
programmes typically already score higher on a number of personality
traits (Openmindedness, Social Initiative, Flexibility, Emotional Stability,
ethnorelativism, international concern, interpersonal communication
skills and self-efficacy) compared to control groups of domestic students
(Leong 2007; Yashima 2010). Environmental sources of these differences
are typically not investigated. However, while controlling for pre-existing
differences, researchers typically find significant increases in the scores of
the volunteers on the various dimensions at the end of the stay abroad.
Immigrants with specific personality profiles may also be more inclined
to engage in interactions with native speakers in their host country.
Ożańska-Ponikwia and Dewaele (2012) found that Polish immigrants
living in English-speaking countries who scored high on Openness and
Self-esteem reported more use of English L2 and that Openness was a
strong predictor of self-perceived English L2 proficiency. In other words,
these personality traits boosted the development of the L2 and increased
the speed of L2 socialisation.
19.3.2 The effect of multilingualism on personality traits

Contrary to cross-psychological research where personality is the predictor
variable, Dewaele and van Oudenhoven (2009) looked at how multilingual-
ism and multiculturalism were linked to personality traits – measured
with the Multicultural Personality Questionnaire. This questionnaire was
developed specifically to measure five personality dimensions that are rele-
vant to multicultural effectiveness (Cultural Empathy, Openmindedness,
Social Initiative, Emotional Stability and Flexibility) (van Oudenhoven,
Timmerman and van der Zee 2007). Participants were ninth-grade pupils at
a Roman Catholic School in Maida Vale, London. Forty-one young immigrant

teenagers5 scored significantly higher than 38 locally born teenagers6 on the

dimensions of Openmindedness and – marginally – on Cultural Empathy,
but they scored significantly lower on Emotional Stability. Teenagers
who reported to be dominant in two languages scored higher on
Openmindedness, marginally higher on Cultural Empathy and significantly
lower on Emotional Stability compared to participants who were dominant
in one language. Functional multilinguals (i.e. users of at least two languages)
scored significantly higher on Cultural Empathy and Openmindedness, and
scored significantly lower on Emotional Stability compared to monolinguals
who were starting to learn a foreign language at school.
Dewaele and Stavans (2014) replicated part of the Dewaele and van
Oudenhoven study in an Israeli context. The main difference was that
the Israeli participants were proficient and frequent users of at least two
languages. The effect of knowing more languages (three to six) had no
effect on scores of the personality dimensions. Israeli-born participants
scored marginally higher on Emotional Stability compared to those born
abroad. Participants with one immigrant parent (but not two) scored
higher on Cultural Empathy, Openmindedness and Social Initiative.
Participants whose language dominance had shifted from their L1 to
Hebrew (either L2, L3 or L4) scored lower than Hebrew L1-dominant parti-
cipants on Emotional Stability. Advanced proficiency and frequent use of
various languages were linked to significantly higher scores on Cultural
Empathy and Openmindedness.
Korzilius, van Hooft, Planken and Hendrix (2011) investigated the link
between language knowledge and personality profile. They considered the
adjustment of international employees of a Dutch multinational company
and used a control group of non-international employees. Those knowing
more foreign languages scored significantly higher on Openmindedness
and Emotional Stability (p. 546). Self-assessed knowledge of foreign
languages was positively linked to Cultural Empathy (p. 546). The more
multilingual international employees scored higher on Openmindedness
and Flexibility than the less multilingual Dutch employees working in the
Netherlands (p. 547), who scored highest on Emotional Stability (p. 549).
Dewaele and Li Wei (2012) investigated the relationship between multi-
lingualism and Cognitive Empathy (Baron-Cohen and Wheelwright 2004)
among 2158 mono-, bi- and multilinguals from around the world. The
authors collected feedback through an online questionnaire, using Likert
scale items. Statistical analysis revealed that the knowledge of more
languages, a bilingual upbringing and the experience of having lived abroad
were, against expectations, not linked to higher levels of Cognitive Empathy.
However, a small but significant positive link emerged between multilingu-
alism (operationalised as high levels of proficiency in several foreign
languages and frequent use of these languages) and Cognitive Empathy.
A follow-up analysis showed that frequent use of multiple languages had a
stronger effect on Cognitive Empathy than mere proficiency in multiple

languages. In other words, highly multilingual participants who use their

languages frequently become more skilful in conversations, as they have
learned to see the world from their interlocutor’s point of view. We inter-
preted this relationship as evidence of multi-competence.
In a follow-up study based on the same database, Dewaele and Li Wei
(2013a) considered the link between multilingualism and Tolerance of
Ambiguity (TA), a lower-order personality trait (Herman et al. 2010).
A significant positive link emerged between the number of languages
known to participants and their TA scores. The monolinguals had the lowest
scores, with the bilinguals in an intermediate position and those with three
or more languages scoring higher. However, the scores of the trilinguals,
quadrilinguals, pentalinguals and sextalinguals were no longer significantly
different. This suggested that the effect of knowing more than three lan-
guages no longer affects the level of TA. A high level of global proficiency in
various languages was also linked to higher TA scores. While growing up bi-
or trilingually from birth had no effect on TA, the experience of having lived
abroad had a strong positive impact, although the effect levelled off after
more than one year abroad. Tolerance of Ambiguity thus appears to be
influenced by an individual’s social–linguistic–cultural environment, and
by the individual’s conscious effort to learn new languages and having to
fit into a new linguistic and cultural environment. When survival in a foreign
environment is at stake, people are forced to attune to local differences. This
brings with it an awareness that their own long-held values, beliefs and
communicative practices are not necessarily shared by their interlocutors.
Moreover, the values, beliefs and communicative practices of the interlocu-
tors may seem baffling, incomplete or contradictory to the L2 user.
Participants who had lived abroad would have had to ‘stretch’ themselves,
manage conflicting cultural, political and ideological perspectives and solve
the paradox of ‘seemingly irreconcilable realities’ (Herman et al. 2010, p. 63).
Dewaele and Li Wei (2013a) showed that a high level of multilingualism
makes individuals more at ease in dealing with ambiguity. However, the
causal pathway could be bidirectional, as it is possible that a higher level of
TA early on in life could strengthen an individual’s inclination to become
multilingual. We argued that the effect of the independent variables on TA
constitute an indication of multi-competence as the presence of various
languages in one mind has effects ‘that go beyond the actual knowledge of
language itself’ (Cook 2002, p. 7).
19.3.3 The effect of multilingualism on lower-order psychological

and affective variables
Research has been carried out linking multilingualism with communi-
cative anxiety (including Foreign Language Anxiety, FLA). This is not a
personality trait per se, but rather a lower-order psychological concept,
or an ‘affective variable’. Results show that participants knowing

more languages typically report lower levels of communicative anxiety

in their various languages, including in their L1 and all the other
languages they know (Dewaele 2007, 2010; Dewaele, Petrides and
Furnham 2008).
The effect of multilingualism has been found to be particularly strong
in several large-scale studies on FLA. Dewaele (2013) found that in the
BEQ database of 1579 multilinguals, collected through a web-based ques-
tionnaire (Dewaele and Pavlenko 2001–2003), the bilinguals reported
consistently higher levels of FLA in their L2 in various situations (speaking
with friends, colleagues, strangers, on the phone and in public) while
pentalinguals reported the lowest levels of FLA in their L2 (p. 184). This
reduced FLA has been linked to the fact that multilinguals have more and
wider experience in communication with a wide range of interlocutors
which has strengthened their ability to avoid or overcome linguistic
icebergs (Dewaele et al. 2008; Dewaele 2013).
Dewaele (2010) has considered the effect of multilingualism on scores
of self-perceived communicative competence and communicative anxiety
in the French of 953 participants who had French as an L1, L2, L3 or L4
extracted from the BEQ. Results showed that knowledge of more
languages in general, and the knowledge of other Romance languages in
particular, were linked to higher self-perceived communicative compe-
tence in French and led to less communicative anxiety using that
language. Participants knowing more languages reported higher levels
of self-perceived competence in speaking, understanding, writing and
reading in French L2, L3 and L4. They also tended to feel less anxious in
speaking with friends, colleagues, strangers, on the phone and in public
in their various languages. Participants who knew more Romance
languages reported feeling more competent in speaking, understanding,
writing and reading in French L2 and L3 and feeling generally less anxious
in speaking with friends, colleagues, strangers, on the phone and in
public in their various languages.
Dewaele and MacIntyre (2014) found that the number of languages that
1746 current foreign language learners from around the world knew was
linked not only to significantly lower levels of FLA but also to signifi-
cantly higher levels of Foreign Language Enjoyment. In other words,
multilingualism lowers negative affect and boosts positive affect in the
foreign language class. Levels of Foreign Language Enjoyment were sig-
nificantly higher than those of FLA, a pattern that was more striking
among more advanced and more multilingual learners and those who
felt their proficiency was above average in their foreign language group.
Finally, three studies focused on the relationship between multilingual-
ism and FLA in homogeneous samples in terms of nationality. Thompson
and Lee (2013) confirmed that the degree of multilingualism of their 123
South Korean college English Foreign Language learners was inversely
linked to their FLA in English. The authors argue that multilingualism in

and of itself has an effect on all language learning experiences. Having

reached a certain level of proficiency in multiple languages has an effect
on FLA levels in different languages: ‘If a learner has multiple experiences
in a language learning environment, it is logical that performing in sub-
sequent language learning environments would be less anxiety inducing’
(p. 17). In a second study on a similar population, Thompson and Lee (2014)
found that very low levels of proficiency in another language did not have
a significant beneficial effect on FLA in the target language. Confirming
the finding in Dewaele and Li Wei (2013), they also found that participants
with more language learning experiences had greater tolerance for ambi-
guity. They explained that having communicated successfully in foreign
language settings, a learner/user understands that not every single lexical
item in a sentence is crucial for overall comprehension (p. 17). Finally,
Thompson (2013) investigated the interface of language aptitude and
multilingualism using a sample of 79 Brazilian language learners of
English. She found that previous language experience had a positive effect
on language aptitude scores, confirming earlier research on the benefits of
bilingualism on L3 acquisition (Sanz 2000). Even participants with a
small amount of previous language learning experience that occurred
post-adolescence, outperformed those with no language learning experi-
ence other than English. Also, multilinguals scored significantly higher
on language aptitude than the bilingual participants. Thompson therefore
claims that the concept of language aptitude is dynamic. Although
she does not refer to the work of Cook, her findings of increased
language aptitude scores could easily be interpreted as evidence of multi-
competence: even small bits of knowledge of various languages gathered
by language learners boost their general cognitive development.
Psychological research has established that increase in certain cognitive
functioning may result in greater creative performance (Kharkhurin 2012,
Chapter 20, this volume). As bilingualism has been linked to increased
cognitive functioning, it is not surprising that studies with bilingual chil-
dren (Landry 1974; Ricciardelli 1992) and college students (Kharkhurin
2008, 2009) showed that these seem to have an edge in divergent thinking,
one of the major components of creativity. These studies also showed
greater divergent thinking performance of participants with high profi-
ciency in both languages as compared to peers with low proficiency in one
language. Kharkhurin (2012) suggests that individuals’ multilingual prac-
tice, or ‘enriched experience’ as Landry put it (1974, p. 10), is linked to the
performance of their language-mediated concept activation, which may
facilitate generative capacity, in other words, simultaneous activation of a
multitude of (un)related concepts. This is facilitated by the multiple links
in the multilingual conceptual system. Kharkhurin (2012) thinks that
language proficiency determines the strength of the connections between
the lexical and conceptual systems in their memory: greater language
proficiency means stronger and more elaborate links to the conceptual

system and more cognitive flexibility. In line with the concept of multi-
competence – which he does not mention – Kharkhurin (2012) argues that
the multilingual’s conceptual system is not merely a combination of two
monolingual ones, and may therefore embed new conceptual representa-
tions in a qualitatively different manner than in a monolingual memory.
Since the conceptual system is shared across both languages, L2 users
would have access to the expanded conceptual representation. The quali-
tatively modified conceptual system could promote the integration of
different, and possibly contradictory concepts, which may increase cogni-
tive flexibility (Kharkhurin 2012).
The benefits of bilingualism for creativity are not always clear-cut.
Kharkhurin (2010) looked at the effect of bilingualism on verbal and non-
verbal criterion-referenced creativity among college students (see also
Kharkhurin, Chapter 20). Russian–English bilinguals with comparable
levels of linguistic proficiency and with similar patterns of language dom-
inance were found to perform better on non-verbal creativity, whereas
monolinguals scored higher on verbal creativity. The bilinguals scored
higher than their monolingual peers on resistance to premature closure,
an important indicator of creativity. In other words, the bilinguals were
less likely to jump to conclusions prematurely. Kharkhurin (2012) argues
that for multilinguals ambiguity is inherent to their linguistic practice,
because the same basic idea may have different nuances in different
languages: ‘This tolerance of ambiguity in turn may facilitate their ability
to keep a pool of possible solutions open long enough to generate a
creative idea’ (p. 118). The greater divergent thinking performance of
multilinguals could be linked to the fact that they perceive the world
through the amalgam of two different conceptual prisms and view events
with a wider range of enriched experiences (Kharkhurin 2012).
Maddux, Adam and Galinsky (2010) investigated the effect of living in
and adapting to foreign cultures on creativity. They found that recalling a
multicultural learning experience facilitated idea flexibility, increased
awareness of underlying connections and associations, and helped over-
come functional fixedness (p. 731). They demonstrated experimentally
that ‘functional learning in a multicultural context . . . is particularly
important for facilitating creativity’ (p. 731). Interestingly, creativity was
found to be enhanced ‘only when participants recalled a functional multi-
cultural learning experience and only when participants had previously
lived abroad’ (p. 731).
Tadmor, Galinsky and Maddux (2012) took this line of investigation one
step further, by looking beyond the effect of mere exposure to new
cultures on creativity and professional success. The authors wondered
why all individuals who had lived abroad for several years did not perform
at the same rate. They argued that while most previous research focused
on shifts in cognitive content as result of exposure to new cultures, it is
important to look also at changes in more general cognitive processes

(p. 521). They found that bicultural individuals outperformed those who
identified more with a single culture. The authors conclude that it is ‘the
simultaneous juxtaposition and synthesis of two cultural perspectives’
that leads to cognitive transformation (p. 537). In other words, ‘although
the living abroad matters, it is how one approaches that experience which
adds critical explanatory value’ (p. 537). The strength of the identification
with both cultures is also crucial: only biculturals with a high level of
identification with both cultures showed greater cognitive and beha-
vioural benefits (p. 537). Although these researchers do not mention
multi-competence, it would be appropriate to evoke the concept, as the
fact of having learnt foreign languages and having lived abroad was more
than the addition of an extra tool for communication and an interesting
set of experiences that could be considered separate, or separable from the
person before the foreign language learning and the stay abroad. What the
results show is that this learning and these experiences changed them in a
fundamental way, not just in their foreign language but also in their L1:
they had become more creative, in other words, they had experienced
general cognitive and behavioural benefits.
19.4 A word of caution
A cautionary note needs to be struck at this point. While significant links

have been reported between multilingualism and psychological variables,
it is important to point out that the effect sizes were always small. In other
words, multilingualism – and the resulting multi-competence – is only one
among many physiological and environmental variables that help shape
our personality.
Another point is that the directionality of the relationship between
multilingualism and psychological variables can never be completely
established. Schrauf (2013) considered the link between bilingual profi-
ciency, and both psychological and social factors. He concluded that the
causal pathway is in fact bidirectional. Indeed, proficiency can be both a
cause and an effect. Certain personality traits and affective dispositions
can strengthen a person’s curiosity and interest in foreign languages.
The resulting action, the learning of a foreign language, or the decision
to live abroad, could then reinforce an initial inclination.
19.5 Conclusion
A growing number of studies have shown that the effect of multilingual-

ism extends beyond the purely cognitive level. Indeed, multilingualism
has been linked to stronger creative behaviour and divergent thinking
(Kharkhurin 2012). A relationship has also been uncovered between

multilingualism and personality traits, with frequent and proficient users

of several languages typically scoring higher on Openmindedness,
Cultural Empathy, Social Initiative (Dewaele and van Oudenhoven 2009;
Korzilius et al. 2011; Dewaele and Stavans 2014), on Tolerance of
Ambiguity (Dewaele and Li Wei 2013a; Thompson and Lee 2013), on
Cognitive Empathy (Dewaele and Li Wei 2012). Moreover, multilinguals
seem to suffer less from FLA (Dewaele 2013; Thompson and Lee 2013,
2014), have more fun in foreign language classes (Dewaele and MacIntyre
2014) and have a better general language aptitude (Thompson 2013).
We argue that all these relationships between multilingualism and
psychological concepts are illustrations of multi-competence, in the
sense that the acquisition of a foreign language ‘alters the L2 user’s mind
in ways that go beyond the actual knowledge of language itself’ (Cook
2002, p. 7). In short, it seems that learning a foreign language tends to
make you a better person, more creative, more openminded, more
empathic, more emotionally stable, more sociable, more likely to enjoy
foreign language classes, better equipped to learn new languages and less
anxious in communication.
Notes
1. A search in the PsycINFO database reveals only 14 hits for ‘multi-

competence’, and all of these refer to the work of applied linguists
(accessed on 22 January 2014).
2. Vivian Cook pointed out (personal communication) that ‘it was an SLA
article in a linguistics-influenced journal’. It might thus be unfair to
blame psychologists for not having noticed it.
3. Cook (2010) pointed out that: ‘The use of “second” should not be taken
too literally. Many sources maintain that it subsumes later languages;
Doughty and Long (2003, p. 3) enumerate how SLA includes “second
(third, etc) languages and dialects”; Lightbown and Spada (2006, p. 204)
say a “second language: . . . may actually refer to the third or fourth
language”.’
4. Ho and Wu (2001) defined cross-cultural psychology as the study
of ‘human behaviour and mental processes, including both their varia-
bility and invariance, under diverse cultural conditions’ (p. 3). See also
Chapter 1 (this volume).
5. They were of African, Arabic, Caucasian and Asian origin, were born
outside the UK, and moved to London. Countries of origin included
Brazil, Italy, Colombia, Congo, The Netherlands, Egypt, Fiji, The
Philippines, Ghana, India, Iran, Iraq, Ireland, Jamaica, Lebanon,
Nigeria, Poland and Portugal.
6. They were of Caucasian and Asian origin but born in the UK.

References
Baron-Cohen, S. and Wheelwright, S. 2004. ‘The empathy quotient: an

investigation of adults with Aspergers syndrome or high functioning
autism, and normal sex differences’, Journal of Autism and Developmental
Disorders 34, 163–175.
Chen, S. X., Benet-Martı́nez, V. and Harris Bond, M. 2008. ‘Bicultural
identity, bilingualism, and psychological adjustment in multicultural
societies: immigration-based and globalization-based acculturation’,
Journal of Personality 76, 803–838.
Davey, G. 2005. Encyclopaedic Dictionary of Psychology. London: Routledge.
Dewaele, J.-M. 2007. ‘The effect of multilingualism, sociobiographical
and situational factors on communicative anxiety and foreign language
anxiety of mature language learners’, The International Journal of
Dewaele, J.-M. 2010. ‘Multilingualism and affordances: variation in self-
perceived communicative competence and communicative anxiety in
French L1, L2, L3 and L4’, International Review of Applied Linguistics 48,
105–129.
Dewaele, J.-M. 2013. Emotions in Multiple Languages, second edition.
Basingstoke: Palgrave Macmillan.
Dewaele, J.-M., Housen, A. and Li Wei 2003. Bilingualism: Beyond Basic
Principles. Clevedon: Multilingual Matters.
competence’, International Journal of Multilingualism 9, 352–366.
Dewaele, J.-M. and Li Wei 2013a. ‘Is multilingualism linked to a higher
tolerance of ambiguity?’ Bilingualism: Language and Cognition 16, 231–240.
Dewaele, J.-M. and Li Wei 2013b. ‘Attitudes towards code-switching among
adult mono- and multilingual language users’, Journal of Multilingual and
Dewaele, J.-M. and Li Wei 2014. ‘Intra- and inter-individual variation in self-
reported code-switching patterns of adult multilinguals’, International
Journal of Multilingualism 11 (2), 225–246

Dewaele, J.-M. and MacIntyre, P. D. 2014. ‘The two faces of Janus? Anxiety
and enjoyment in the Foreign Language Classroom’, Studies in Second
Language Learning and Teaching 4 (2), 237–274.
Dewaele, J.-M. and Pavlenko, A. 2001–2003. Web questionnaire:
Bilingualism and Emotions. University of London.
Dewaele, J.-M., Petrides, K. V. and Furnham, A. 2008. ‘The effects of trait
emotional intelligence and sociobiographical variables on communica-
tive anxiety and foreign language anxiety among adult multilinguals: a
review and empirical investigation’, Language Learning 58, 911–960.
Dewaele, J.-M. and Stavans, A. 2014. ‘The effect of immigration, accultura-
tion and multi-competence on personality profiles of Israeli multilin-
guals’, The International Journal of Bilingualism 18 (3), 203–221.
Dewaele, J.-M. and van Oudenhoven, J. P. 2009. ‘The effect of multilingual-
ism/multiculturalism on personality: no gain without pain for third
culture kids?’, International Journal of Multilingualism 6, 443–459.
Doughty, C. and Long, M. 2003. Handbook of Second Language Acquisition.
Oxford: Blackwell.
Grosjean, F. 1989. ‘Neurolinguistics, beware! The bilingual is not two
monolinguals in one person’, Brain and Language 36, 3–15.
Grosjean, F. 2008. Studying Bilinguals. Oxford: Oxford University Press.
Hammarberg, B. (ed.) 2009. Processes in Third Language Acquisition.
Edinburgh: Edinburgh University Press.
Herman, J. L., Stevens, M. J., Bird, A., Mendenhall, M. and Oddou, G. 2010.
‘The Tolerance for Ambiguity Scale: towards a more refined measure for
international management research’, International Journal of Intercultural
Relations 34, 58–65.
Ho, D. Y. F. and Wu, M. 2001. ‘Introduction to cross-cultural psychology’, in
L. L. Adler and U. P. Gielen (eds), Cross-cultural Topics in Psychology, pp. 3–13.
Westport, CT: Praeger.
Kharkhurin, A. V. 2008. ‘The effect of linguistic proficiency, age of second
language acquisition, and length of exposure to a new cultural environ-
ment on bilinguals’ divergent thinking’, Bilingualism: Language and
Kharkhurin, A. V. 2009. ‘The role of bilingualism in creative performance
on divergent thinking and Invented Alien Creatures tests’, Journal of
Creative Behavior 43, 59–71.
Kharkhurin, A. V. 2010. ‘Bilingual verbal and nonverbal creative behavior’,
International Journal of Bilingualism 14, 1–16.
Kharkhurin, A. V. 2012. Multilingualism and Creativity. Bristol: Multilingual
Matters.
Kihlstrom, J. F. 2013. ‘The person–situation interaction’, in D. Carlston
(ed.), Oxford Handbook of Social Cognition, pp. 786–805. Oxford University
Press.
Kim, Y. Y. 2001. Becoming Intercultural: An Integrative Theory of Communication
and Cross-cultural Adaptation. Thousand Oaks, CA: Sage Publications.

Korzilius, H., van Hooft, A., Planken, B. and Hendrix, C. 2011. ‘Birds of
different feathers? The relationship between multicultural personality
dimensions and foreign language mastery in business professionals
working in a Dutch agricultural multinational’, International Journal of
Intercultural Relations 35, 540–553.
Landry, R. G. 1974. ‘A comparison of second language learners and mono-
linguals on divergent thinking tasks at the elementary school level’.
Modern Language Journal 58, 10–15.
Leong, C. H. 2007. ‘Predictive validity of the Multicultural Personality
Questionnaire: a longitudinal study on the socio-psychological adapta-
tion and intercultural experience among Asian graduates’, International
Journal of Intercultural Relations 31, 545–559.
Lightbown, P. and Spada, N. 2006. How Languages are Learned. Oxford
University Press.
Maddux, W. W., Adam, H. and Galinsky, A. D. 2010. ‘When in Rome . . .
learn why the Romans do what they do: how multicultural learning
experiences enhance creativity’, Personality and Social Psychology Bulletin
3, 731–741.
Magnusson, D. and Endler, N. S. (eds) 1977. Personality at the Crossroads:
Current Issues in Interactional Psychology. Hillsdale, NJ: Erlbaum.
McCrae, R., Costa Jr, P. T., Ostendorf, F., Angleitner, A., Hrebı́cková,
M., Avia, M. D., Sanz, J., Sánchez-Bernardos, M. L., Kusdil,
M. E., Woodfield, R., Saunders, P. R. and Smith, P. B. 2000. ‘Nature over
nurture: temperament, personality, and life span development’, Journal
of Personality and Social Psychology 78, 173–186.
Mitchell, R. and Myles, F. 1998. Second Language Learning Theories. London:
Edward Arnold.
Ożańska-Ponikwia, K. and Dewaele, J.-M. 2012. ‘Personality and L2 use:
openminded and self-confident individuals are more likely to be active
and proficient L2 users’, EUROSLA Yearbook 12, 112–134.
Pervin, L. A. and Cervone, D. 2010. Personality: Theory and Research, 11th
edition. Hoboken, NJ: Wiley.
Petrides, K. V., Pita, R. and Kokkinaki, F. 2007. ‘The location of trait
emotional intelligence in personality factor space’, British Journal of
Ricciardelli, L. A. 1992. ‘Bilingualism and cognitive development in
relation to threshold theory’, Journal of Psycholinguistic Research 21,
301–316.
Sanz, C. 2000. ‘Bilingual education enhances third language acquisition:
evidence from Catalonia’, Applied Psycholinguistics 21, 23–44.
Schrauf, R. 2013. ‘Using correspondence analysis to model immigrant
multilingualism over time’, in J. Duarte and I. Gogolin (eds), Linguistic
Super-diversity in Urban Areas: Research Approaches, pp. 27–44. Amsterdam:
John Benjamins.

Tadmor, C. T. Galinsky, A. D. and Maddux, W. W. 2012. ‘Getting the most

out of living abroad: biculturalism and integrative complexity as key
drivers of creative and professional success’, Journal of Personality and
Social Psychology 103, 520–542.
Thompson, A. 2013. ‘The interface of language aptitude and multilingual-
ism: reconsidering the bilingual/multilingual dichotomy’, Modern
Language Journal 97, 685–701.
Thompson, A. S. and Lee, J. 2013. ‘Anxiety and EFL: does multilingualism
matter?’, International Journal of Bilingual Education and Bilingualism 16,
730–749.
Thompson, A. S. and Lee, J. 2014. ‘The impact of experience abroad and
language proficiency on language learning anxiety’, TESOL Quarterly 48,
252–274.
Twain, M 1869. Innocents Abroad. Hartford, CT: American Publishing
Company.
Van Oudenhoven, J. P., Timmerman, M. and van der Zee, K. 2007. ‘Cross-
cultural equivalence and validity of the Multicultural Personality
Questionnaire in an intercultural context’, Journal of International
Communication 13 (2), 51–65.
Wang, X. L. 2008. Growing up with Three Languages. Clevedon: Multilingual
Matters.
Yashima, T. 2010. ‘The effects of international volunteer work experiences
on intercultural competence of Japanese youth’, International Journal of
Intercultural Relations 34, 268–282.

20
Multi-competence as a
creative act: ramifications
of the multi-competence
paradigm for creativity
research and creativity-
fostering education
Anatoliy V. Kharkhurin
20.1 Introduction
Empirical investigation of multilingual creativity seems to tap, not into the

multilingual mind, but into a race between a group of people classified as
bilinguals and those classified as monolinguals. It appears to be a sort of all-
round event in which representatives of these groups are being systemati-
cally compared on a variety of creativity tests. These tests were developed in
the psychometric tradition perceiving creative capacity as an ability to
initiate multiple cycles of divergent and convergent thinking (Guilford
1967). A combined effort of these two types of thinking creates an active,
attention-demanding process that allows generation of ideas satisfying the
defining characteristics of a creative product: novelty (i.e. original or unex-
pected) and utility (i.e. useful or meeting task constraints);1 see Sternberg
(1999) for an overview.2 Over the last half-century, numerous studies have
provided evidence for the ability of divergent thinking tests to predict certain
aspects of creative problem-solving performance and real-world creative
achievement. Although, as Runco (1991) argued, “Divergent thinking is not
synonymous with creative thinking” (p. ix), many researchers believe that
divergent thinking is a defining component of the creative process (Lubart
2000). Guilford associated the properties of divergent thinking with four
main characteristics: fluency (the ability to rapidly produce a large number

Multi-competence as a creative act 421
of ideas or solutions to a problem), flexibility (the capacity to consider a

variety of approaches to a problem simultaneously), elaboration (the ability
to think through the details of an idea and carry it out), and originality (the
tendency to produce ideas different from those of most other people).
Thus, bilinguals were found to systematically outperform their monolin-
gual counterparts on divergent thinking tests. Kharkhurin (2012) presented a
list of studies investigating this issue. These studies demonstrated bilinguals’
advantages on divergent thinking traits such as fluency (e.g. Carringer 1974;
Hommel, Colzato, Fischer and Christoffels 2011; Jacobs and Pierce 1966;
Karapetsas and Andreou 1999; Ricciardelli 1992a), flexibility (e.g. Adi-Japha,
Berberich-Artzi and Libnawi 2010; Carringer 1974; Konaka 1997), elaboration
(e.g. Kharkhurin 2008; Srivastava and Khatoon 1980; Torrance, Gowan, Wu
and Aliotti 1970), and originality (e.g. Cummins and Gulutsan 1974;
Kharkhurin 2009; Konaka 1997; Okoh 1980); on insight problems (Cushen
and Wiley 2011); and on structured imagination tasks (Kharkhurin 2009).
Then researchers finished in glorious celebration of bilinguals’ superior
abilities. Indeed, what else can one add to this exhaustive conclusion?
At the same time, this conclusion raises two obvious questions. First,
real-life observations do not support empirical findings demonstrating
greater creative performance of individuals speaking multiple languages.
For example, the European Commission’s (2012) survey “Europeans and
their languages” reported frequency of use of a first language different
from the mother tongue for 27 EU member states. The data were compared
with the creative class of these countries (Florida 2002), which represents
the portion of a population that is involved in creative work. The creative
class consists of “workers in fields spanning science and technology, busi-
ness and management, healthcare and education, and arts, culture, and
entertainment” (Florida, Mellander and Stolarick 2011, p. 29).
Surprisingly, the correlation between the frequency of second language
use and the creative class barely exceeded zero. A possible explanation for
this discrepancy stems from a complexity of the creativity construct.
Creativity is perceived as a syndrome or complex (Runco 2014) prompted
by a large variety of factors such as education, expertise, motivation,
attitudes, personality traits, personal experience, and socioeconomic and
sociocultural conditions. Multilingualism may play an insignificant role
here, and its effect can be overridden by those factors. In other words, the
specific economic, political, social, cultural, and educational aspects of an
individual’s development may have an impact on his or her creative
performance above and beyond the effect of multilingualism.
Nevertheless, we find a consistent pattern of bilingual advantages in
empirical research. This suggests that these studies may overlook some
important determinants of creative performance. Thus, the second ques-
tion calls for systematic investigation of specific cognitive mechanisms
facilitating creative performance, which may be encouraged by multilin-
gual practice. The answer to this question, however, cannot be obtained

422 A N ATO L I Y V. K H A R K H U R I N
from a comparison of bilinguals and monolinguals, since people in these

two groups seem to have two distinct cognitive structures (see the discus-
sion below).
This idea directly refers to the theme of the present volume. Speaking
multiple languages adds not only to an individual’s linguistic repertoire:
multilingual practice influences one’s cognitive functions, conceptual
representations, and even personality traits (see Chapter 1 for a discus-
sion). Under this assumption, multilinguals and monolinguals appear to
possess distinctively different minds and therefore their comparison
would not shed any light on the creative advantages of the former. Thus,
a body of empirical data accumulated in research on bilingual creativity
would not illuminate the problem in question. Rather, we need to
systematically investigate multilingual speakers with different histories
of language acquisition and thus identify which factors can facilitate
cognitive processes underlying creativity.
Hence, instead of providing a rationale for introducing the multi-com-
petence perspective and defending it against the monolingual perspective,
which is extensively presented in many chapters of this volume, this
chapter employs the framework of multi-competence to present the next
step in research on multilingual creativity. The focus group in this
research should consist of individuals with different histories of acquisi-
tion and use of their languages. This ensures manipulation of various
factors related to an individual’s multi-competence. These factors should
be carefully examined with respect to their potential impact on cognitive
mechanisms and personality traits contributing to an increase in the
individual’s creative abilities. Once the link between an individual’s
multi-competence and creative functioning is established, researchers
can think about methods and strategies facilitating these connections
and therefore increasing the individual’s multi-competence and creative
potentials. That is, education researchers and educators should look into
development of educational programs fostering these instances.
The following sections elaborate on each of these points and present
evidence supporting this ground plan. Section 20.2 presents empirical
evidence relating various aspects of multi-competence to individuals’
creative performance. Sections 20.3–20.5 present evidence for cognitive
mechanisms and personality traits facilitated by those aspects. Finally,
Section 20.6 lays the theoretical groundwork for an educational program
fostering multi-competence and creativity.
20.2 Aspects of multi-competence influencing

creative performance
Language proficiency appears to be the first multi-competence aspect that

might have an impact on creative performance. Indirect evidence for the

role of language proficiency in an individual’s creative abilities comes

from studies comparing bilinguals with monolinguals. Cummins’ (1976)
threshold theory predicted that bilinguals need to achieve high levels of
proficiency in both of their languages before bilingualism can promote
cognitive advantages. This was demonstrated in studies with bilingual
children and college students. They revealed greater performance in diver-
gent thinking among participants with high proficiency in both languages
as compared to their linguistically unbalanced counterparts. For example,
Ricciardelli’s (1992a) study with Italian–English bilingual and English
monolingual children revealed that only bilinguals with high proficiency
in both languages showed greater fluency and imagination compared with
their monolingual counterparts. Similarly, Lemmon and Goggin’s (1989)
study with Spanish–English bilingual college students found that only
those with low language proficiency performed less well than their mono-
lingual counterparts on fluency and flexibility. In the same fashion,
Konaka (1997) reported that the degree of bilingual balance (based on the
score computed from Japanese–English bilingual children’s self-rating and
their performance on the Word Association Test: Lambert 1956) signifi-
cantly predicted performance on fluency, flexibility, and originality.
A few recent studies conducted with different types of bilinguals in
various geographic locations and cultural contexts have confirmed the
effect of language proficiency on an individual’s creative capacities.
Bilinguals with high proficiency in both English and Russian performed
better on elaboration than their less proficient counterparts (Kharkhurin
2008). Similarly, Farsi–English bilinguals who were highly proficient in
both languages outperformed their unbalanced and moderately proficient
counterparts on fluency (Kharkhurin 2009). Moreover, Lee and Kim (2011)
found that more balanced Korean–English bilinguals (whose level of bilin-
gualism was assessed by the aforementioned Word Association Test)
obtained higher creativity scores than their less balanced counterparts.
These findings were complemented by those from another study con-
ducted with bilinguals with different proficiency levels in English
(Kharkhurin 2011). This study revealed that more linguistically proficient
bilinguals tended to achieve higher scores on originality and revealed
more unstructured imagination (as measured by the Invented Alien
Creature test: see Ward 1994).
The second multi-competence aspect refers to the age at which indivi-
duals acquired their languages. Traditionally, a distinction is made
between simultaneous and sequential bilinguals (McLaughlin 1984).
Simultaneous bilinguals learn both of their languages from the onset of
language acquisition. Sequential bilinguals learn their second language
(L2) after the age of five, when the basic components of the first language
(L1) are already in place. Sequential bilinguals are further divided into
early and late, reflecting the age at which L2 acquisition occurred
(Genesee 1978). A group of simultaneous Armenian–Russian bilinguals

achieved higher scores on flexibility and originality than their sequential

counterparts who started learning one of the two languages two to four
years later (Kostandyan and Ledovaya 2013). There is also evidence that
Russian–English bilinguals who acquired their L2 at a younger age
achieved higher scores on fluency and flexibility (Kharkhurin 2008).
Similarly, bilinguals who acquired their L2 (English) by the age of six
tended to solve insight problems more readily than their counterparts
who acquired L2 after this age (Cushen and Wiley 2011).
The third multi-competence aspect addresses a defining feature of multi-
lingualism, namely code-switching – the alternation and mixing of differ-
ent languages in the same episode of speech production. Code-switching
has been argued to be a creative act (e.g. Li Wei and Wu 2009). For example,
linguists working in the linguistic ethnography tradition have replaced
code-switching with other terms such as translanguaging to capture its
creative and dynamic nature (see the review in Garcia and Li Wei 2014
and Chapter 26 of this volume). They investigated the use of translangua-
ging in diverse contexts, from literature and drama, to pop songs, the new
media, and public signs (e.g. Androutsopoulos 2015; Chik 2010; Jonsson
2005; Sebba, Jonsson and Mahootian 2012). These studies took code-
switching not simply as a juxtaposition of different grammatical structural
elements, but as an expressive and creative performance. Bhatia and
Ritchie (2008) revealed various facets of bilingual creativity through
code-switching as it manifests itself in the day-to-day verbal behaviour of
a bilingual. They argued that code-switching is essentially an “optimizing”
strategy, rendering a wide variety of new meanings which the separate
linguistic systems are incapable of rendering by themselves. The only
empirical evidence for the relationship between code-switching and crea-
tivity demonstrated that those individuals who code-switch frequently
regularly obtained higher scores on originality than those who do not
code-switch in their everyday practice (Kharkhurin and Li Wei 2014).
The fourth multi-competence aspect reflects the context of language
acquisition and use. Studies of bilingual creativity have generally disre-
garded the fact that most participants in the target samples experience and
participate in more than one culture. These individuals are primarily
immigrants, migrant workers, members of the minority groups, or foreign
students exposed to different educational systems. They acquire each of
their languages in the respective cultural environments where different
cultural cues are available (Pavlenko 2000). Therefore, in addition to
acquiring several languages, they could also adopt a range of multicultural
values and beliefs. Acculturation studies support this view by demonstrat-
ing that language acquisition is often accompanied by adoption of the
cultural values of the country in which this language is acquired (e.g.
Birman, Trickett and Vinokurov 2002; Gordon 1964). On the other hand,
research in creativity has demonstrated that the specific economic, poli-
tical, social, and cultural aspects of the environment can have a

considerable influence both on levels of creative potential and on how

creativity is evaluated (e.g. Lubart 1999). Sociocultural values and norms
determine and shape the concept of creativity, which may in turn influ-
ence the manner in which creative potential is apprehended and incar-
nated. Thus, if multilinguals acquire their languages in different countries,
they are most likely to have been exposed to different sociocultural
environments. This multicultural experience encourages variation in the
development of creative potential. Therefore, multilingual individuals’
experience with multiple sociocultural settings may increase their crea-
tive potential.
This argument finds support in cross-cultural research demonstrating
that the effect of multilingualism on creative performance is often con-
founded with the effect of multiculturalism (see Kharkhurin 2012 for a
discussion). For example, Kharkhurin (2008) found that the length of
residence in the new cultural environment was related to Russian–
English bilingual college students’ fluency, flexibility, and elaboration
above and beyond the effect of bilingualism. Similar findings were
obtained by Maddux and Galinsky (2009), who found that the amount of
time MBA students from 40 different nations had lived abroad significantly
predicted creative solutions of Duncker’s (1945) candle-mounting problem
when the effect of bilingualism was controlled.
Another line of research proposes that, in addition to direct contribution
to this performance, the specific settings of the sociocultural environment
to which an individual was exposed may modulate the impact of multi-
lingualism on creativity (e.g. Kharkhurin 2010; Leung, Maddux, Galinsky
and Chiu 2008). This idea stems from cross-cultural research in creativity
demonstrating that variations in the manner of socialisation, degrees of
self-perception and self-expression, and education and social conduct may
modulate the differences in creative performance of the representatives of
different cultures (e.g. Kharkhurin and Samadpour Motalleebi 2008; Niu
and Sternberg 2001; Zha et al. 2006). If individuals’ creative potential may
be influenced by their experience with different cultures, the variations in
multilinguals’ cultural settings may have an impact on different aspects of
their creative thinking. For example, Kharkhurin (2010) compared Farsi–
English bilingual and Farsi monolingual college students residing in the
Middle East with their Russian–English bilingual and English monolingual
counterparts residing in the USA. The study demonstrated that the inter-
action between bilingualism and the sociocultural environment had a
significant influence on creative performance. Moreover, this study specu-
lated that the cultural distance between the environments to which bilin-
gual groups were exposed in the respective countries could also play a role
in an individual’s creative behavior.
The fifth multi-competence aspect also addresses the context of lan-
guage acquisition and use. However, this one is concerned with emotional
experience. The literature on creativity regards an individual’s emotional

state as an important determinant and a powerful engine of creative

behavior (e.g. Averill, Chon and Hahn 2001; Kaufmann 2003). Various
researchers have claimed that emotions can either help or hurt creative
endeavors (e.g. Montgomery, Hodges and Kaufman 2004; Russ 1999). At the
same time, an extensive literature demonstrates that bilinguals’ emotions
are realized differently in their different languages and that bilingual
individuals have particular preferences for one or another language
when it comes to expression of their emotions (see Pavlenko 2006 for
an overview). Therefore, the experience of different emotions in different
linguistic contexts can lead to variation in creative performance.
Kharkhurin and Altarriba (2015) tested this thesis with Arabic–English
bilingual college students who were induced into positive or negative
mood states using either an Arabic or an English linguistic context. The
study revealed an interactive effect of the emotional state and linguistic
context on an individual’s creativity. Specifically, participants obtained
significantly greater non-verbal originality when they were induced into
a positive mood state while being tested in English or a negative mood
state while being tested in Arabic, as compared to when they were
induced into a positive mood state while being tested in Arabic or a
negative mood state while being tested in English. The role of emotions
in multilingual creativity was hinted at in Kharkhurin and Li Wei’s (2014)
study, in which emotion-triggered code-switching was reported to link to
creative capacities. Specifically, code-switching induced by a particular
emotional state appeared to relate to an increase in originality in think-
ing. Recent research on the relationship between multilingualism and
emotions has provided evidence that code-switching is often triggered by
the speaker’s emotions and other affective factors (e.g. Dewaele 2010;
Pavlenko 2005). This, in turn, could boost one’s creativity, possibly
because code-switching breaks the conventional conversational routine
by introducing different, emotionally laden elements of language. Think,
for example, about cognitive poetry, in which two or more languages are
mixed together to establish a textual plane with multiple emotionally
laden connotations.
20.3 Cognitive mechanisms underlying multilingual

creativity
So far we have identified five multi-competence aspects which might con-

tribute to multilingual creativity: language proficiency, age of language
acquisition, code-switching, cross-cultural, and emotional experiences.
The second goal of this chapter is to identify cognitive mechanisms under-
lying creativity and how those aspects can facilitate them.
Recall that in the psychometric literature, creative thinking is perceived
as an ability to initiate multiple cycles of divergent and convergent

thinking (Guilford 1967). The fundamental difference between these two

processes is that convergent thinking is a conscious attention-demanding
process, whereas divergent thinking occurs in an unconscious mind where
attention is defocused (e.g. Kasof 1997; Mendelsohn 1976) and thought is
associative (e.g. Koestler 1964; Mednick 1962; Ward, Smith and Vaid 1997).
Convergent thinking seeks one correct answer to a question or solution to
a problem, which must have a single answer or solution (Runco, Dow and
Smith 2006). Divergent thinking, on the other hand, involves a broad
search for information and generation of numerous novel alternative
answers or solutions to a problem which has no single solution (Guilford
1967). The solutions generated during divergent thinking are subsequently
evaluated during convergent thinking, which narrows all possible alter-
natives down to a single solution.
There is evidence that multilingual practice facilitates both divergent
and convergent thinking. The former may benefit from the specific archi-
tecture of multilingual memory, and the latter may benefit from multi-
linguals’ highly developed selective attention.
The functioning of divergent thinking can be explained by an automatic
spreading activation mechanism that simultaneously triggers a large num-
ber of mental representations. These representations are stored in con-
ceptual memory. The latter is assumed as a pattern of spreading activation
(McClelland and Rumelhart 1985) over a large set of mutually linked units
of meaning (or conceptual features) organized in conceptual networks
(Lamb 1999). The spreading activation mechanism transfers activation
between conceptual features, providing facilitation for related concepts
and inhibition for unrelated ones. This property of the conceptual system
has been illustrated in priming studies (e.g. Meyer and Schvaneveldt 1971)
that show that semantically related words tend to influence each other.
The associations between distant mental representations can be estab-
lished because of the distributed nature of the conceptual system (see
Kharkhurin 2012 for details). In light of this discussion, divergent thinking
takes place when a large number of conceptual representations are
accessed simultaneously. Spreading activation among distributed concep-
tual representations may build the links between distant, often unrelated
ideas. Simultaneous activation of these ideas may establish a rich plane of
thought from which original and novel ideas might be extracted.
The specific architecture of multilingual memory is argued to facilitate
greater spreading activation between conceptual representations and
thereby stimulate divergent thinking (see Kharkhurin 2012 for a detailed
discussion). This may be accomplished through language-mediated con-
cept activation (LMCA) (Kharkhurin 2007, 2008). This mechanism is based
on the assumption that translation equivalents automatically activate
each other through shared conceptual representations (e.g. concept-
mediated translation in Kroll and de Groot 1997). Although translation
equivalents share most of the conceptual features, these representations

are not identical (e.g. Paradis 1997). Variations in the conceptual represen-
tations of translation equivalents may result in the simultaneous activa-
tion of additional concepts, which may eventually produce a large
pattern of activation over unrelated concepts from different categories.
The activation of these concepts is assumed to take place through the
lemmas representing the translation equivalents in multiple languages
and/or through the word forms (e.g. phonetic, orthographic) shared by
these languages. The evidence for LMCA is provided by cross-language
studies using semantic (see Kroll and Tokowicz 2005 for a review) and
translation (see Altarriba and Basnight-Brown 2007 for a review) prim-
ing paradigms. These studies revealed that automatic spreading activa-
tion takes place not only between translation equivalents in different
languages, but between semantically related words in different lan-
guages as well. The elaborate LMCA may allow multilinguals to process
a large number of unrelated concepts from different categories simulta-
neously; that is, it may stimulate multilinguals’ divergent thinking. The
empirical evidence for LMCA was obtained in a study comparing the
performance of Russian–English bilingual and Russian monolingual
college students on divergent thinking and a cross-language priming
test (Kharkhurin and Isaeva 2015). The latter presented participants
with pairs of Russian words, which were unrelated in Russian but
related through their translation equivalents in English (e.g. филиал
[branch] is unrelated to дерево [tree], but “branch” is related to “tree”).
Bilingual participants showed stronger priming effect and revealed
greater flexibility in thinking compared with their monolingual coun-
terparts. These results demonstrated that bilinguals’ divergent thinking
can benefit from elaborate LMCA.
The purpose of convergent thinking is to find the single best (or correct)
answer to a clearly defined problem (Cropley 2006). This cognitive func-
tion appears inevitable when a large pool of ambiguous divergent
thoughts needs to be narrowed down to a single creative solution. These
possible candidates should be explored, criticized, and evaluated to select
the best fit to the problem. Kharkhurin (2011) argued that individuals’
efficient selective attention may support creative problem-solving at the
stage where a conscious attention-demanding process assists in narrowing
a multitude of possible alternatives down to a single original solution. At
the same time, Bialystok and her colleagues demonstrated that selective
attention is enhanced by bilingual practice (review in Bialystok 2005).
They argued that bilinguals’ selective attention is facilitated by their
extensive practice with two active language systems, during which they
have to solve the conflicts in lexical retrieval, which are unraveled by
efficient executive control. Kharkhurin identified two control mechan-
isms of selective attention that may contribute to the improvement of
bilinguals’ creative abilities. The facilitation of relevant information was
likely to boost the ability to activate a multitude of unrelated concepts and

to work through the concepts that are already activated. The inhibition of
irrelevant information seemed to enhance the capacity to produce original
and useful ideas.
20.4 Personality traits underlying multilingual creativity
In addition to influence on cognitive mechanisms underlying creative

thinking, multilingual practice seems to have an impact on personality
traits associated with creative behavior.
Cognitive flexibility is considered an important creativity trait (Guilford
1968). This trait allows an individual to find different perspectives, to
switch between perspectives, and to think outside the box. Flexibility
in thinking may be highly suitable in situations requiring non-trivial
manipulation of existing cognitive structures, especially when logical
approaches fail to produce satisfactory results (Torrance and Safter
1999). Recall that flexibility is included in Guilford’s (1967) characteristics
of divergent thinking. Bilingual individuals were found to outperform
their monolingual counterparts on this trait (e.g. Carringer 1974;
Kharkhurin 2008; Konaka 1997). In addition, Kharkhurin (2009) reported
the effect of bilingualism on structured imagination. The latter limits
individuals’ thinking outside the box; that is, people have difficulties
violating the conceptual boundaries of a standard category when creating
a new exemplar of that category. The ability to violate these boundaries
can be ascribed to cognitive flexibility as well. Farsi–English bilinguals
violated the conceptual boundaries more readily than their monolingual
counterparts. Moreover, current research has shown that both bilingual
children and adults excel in non-linguistic tasks requiring cognitive flex-
ibility (e.g. Adi-Japha et al. 2010; Carlson and Meltzoff 2008; Costa,
Hernandez and Sebastián-Gallés 2008). These studies focus on multilin-
guals’ executive control functions and therefore hint at the possibility that
cognitive flexibility builds on inhibitory control mechanisms, such as
selective attention, mentioned above.
Tolerance of ambiguity appears to be another personality trait
related to an individual’s creative behavior (e.g. Zenasni, Besançon
and Lubart 2008). Budner (1962) defined this trait as the “tendency to
perceive ambiguous situations as desirable” (p. 29). In a search for a
creative answer, people may face a situation in which they have to
consider several contradictory ideas simultaneously. People perceive
these situations as ambiguous and tend to avoid them by leaping to
the first available solution. However, those who tolerate ambiguity may
be able to keep a pool of possible solutions open long enough to work
effectively on a larger set of ideas, which may eventually result in
generating a creative idea. Speaking multiple languages was found to
relate to tolerance of ambiguity (Dewaele and Li Wei 2013). In this

study, individuals speaking more than two languages revealed greater

tolerance of ambiguity in comparison with those speaking two or only
one language. Second language acquisition researchers also identified
this trait as an important contributor to successful L2 learning (e.g.
Oxford and Ehrman 1992; Rubin 1975).
Further, recent research has provided evidence that advanced knowledge
and frequent use of more languages is linked to open-mindedness (Dewaele
and Stavans 2014; Dewaele and van Oudenhoven 2009). The operational
definition of this trait refers to openness to different sociocultural con-
structs: “open and unprejudiced attitude towards outgroup members and
towards different cultural norms and values” (Dewaele and van Oudenhoven
2009, p. 449). By analogy, one can think about openness to new ideas and
experiences, which has been always regarded as an important creativity trait
(e.g. Feist 1998; Silvia et al. 2009). People with high scores on openness to
experience are artistic, inventive, and curious; they appreciate novel ideas
and new experiences, consider a variety of perspectives and thoughts, and
employ unconventional problem-solving strategies (McCrae and Costa Jr
1997). There is no empirical evidence of a link between multilingual practice
and openness to experience. Researchers who might become interested in
this theme should look at the role of a multilingual individual’s multicul-
tural experience, which is discussed in the following section.
Finally, motivation has been shown to play a crucial role in both success-
ful language learning (e.g. Engin 2009; Masgoret and Gardner 2003; Wang
2008) and prolific creative behavior (e.g. Amabile 1996; Hennessey 2010).
Current theories considering the role of motivation in creativity (see
Collins and Amabile 1999 for an overview) emphasize that creativity gen-
erally prospers under conditions that support intrinsic motivation (stimu-
lated by personal interest and inner potential) and suffocates under
conditions accentuating extrinsic motivation (such as rewards and incen-
tives). Therefore, one may expect that individuals with strong intrinsic
motivation may succeed in both language learning and creative perfor-
mance. For example, Kharkhurin (2010) claimed that intrinsic motivation
could mediate the effect of bilingualism on creativity. He based this argu-
ment on a speculative interpretation of his findings that only those bilin-
gual participants who were intrinsically motivated to change their country
of residence showed greater originality in thinking.
20.5 Linking multi-competence aspects with cognitive

mechanisms and personality traits underlying
creative thinking
Once the multi-competence aspects, cognitive mechanisms, and person-

ality traits contributing to multilingual creativity have been identified, we
can proceed to establish the links between them.

Proficiency in the languages used by an individual and the age at which

these languages were acquired could shape the LMCA. The proficiency in
respective languages determines the strength of the connections between
the lexical (lemmas and word forms: see discussion above) and conceptual
systems in multilingual memory. The degree of linguistic skills may influ-
ence the intensity of the lexical access: greater language proficiency may
result in establishing stronger and more elaborate links to the conceptual
system. As a result, more concepts become readily available for the LMCA.
Therefore, individuals who attained a high expertise in their languages
would have stronger and more efficient links between lexical and concep-
tual levels than those who were not able to develop any of their languages
to a high degree. As a result, they would employ the LMCA mechanism
more effectively.
Further, the studies examining the neural underpinnings of multilin-
gual representations (e.g. Dehaene et al. 1997; Kim, Relkin, Lee and
Hirsch 1997) suggested that the age of language acquisition has an impact
on the structure and/or functioning in the multilingual system.
Individuals who acquired their languages early in life may develop a
greater sensitivity to underlying concepts and more refined connections
between lexical and conceptual representations for their languages. The
memory system in early multilinguals might foster the LMCA by provid-
ing fast routing of information exchange between both lexicons and
concepts. On the other hand, late multilinguals may develop an asym-
metry in lexical access (see Kroll and de Groot 1997 for detailed discus-
sion), with fewer shared conceptual features having direct links from
respective lexicons. The asymmetrical routing in the lexical–conceptual
network may result in a less efficient LMCA. Thus, due to a more symme-
trical distribution of lexical–conceptual routes, more conceptual repre-
sentations become readily available for the LMCA in individuals who
acquired their languages early in life.
The degree of efficient employment of selective attention may be
stipulated by language proficiency, age of language acquisition, and
intensity of code-switching. Kharkhurin (2011) found that individuals
with different levels of language proficiency seem to employ different
strategies of creative problem-solving based on which mechanisms of
selective attention they utilize. Highly proficient bilinguals are likely to
exercise the inhibition mechanism to suppress interference of one lan-
guage during processing of the other (Bialystok 2001; Green 1998).
Suppression of rudimentary solutions in creative problem-solving may
facilitate the extraction of an original solution. Inhibition of typical
characteristics of a category may result in focusing attention on uncon-
ventional, atypical solutions to creative problems. On the other hand, in
light of an assumption that, during early stages of L2 acquisition, learners
access the meaning of L2 words through their L1 (see the discussion of the
Revised Hierarchical Model of bilingual memory in Kroll and Tokowicz

2005), bilinguals with moderate language proficiency are likely to exer-

cise the facilitation mechanism to maintain the links between the two
languages. This mechanism may encourage greater activation of unre-
lated concepts, which spreads the activation to the conceptual represen-
tations lying beyond a standard set of category properties. This allows
bilinguals to simultaneously consider a large number of properties from
different categories, and therefore to overcome the limitations of struc-
tured imagination (see discussion above).
Further, early language acquisition was argued to improve executive
control functions. This evidence comes from the studies comparing early
bilingual children and their monolingual counterparts, which demon-
strated the advantages of the former on a number of executive control
tasks including dimensional-change card-sorting tasks (Bialystok and
Martin 2004) and identifying and shifting between multiple interpreta-
tions of ambiguous-figure and figure–ground illusions (Bialystok and
Shapero 2005). The older adults who were bilingual from an early age
demonstrated similar advantages in executive function over age-matched
monolinguals (e.g. Bialystok, Craik, Klein and Viswanathan 2004).
Moreover, Bialystok (2001) argued that L2 acquisition later in life, even
to the point of balanced fluency, may not yield the same cognitive advan-
tages. Considering that selective attention appears to be one of the execu-
tive control functions, one may expect that selective attention is also
stipulated by the age of language acquisition.
Active code-switching may also provide extensive opportunities for
selective attention. Based on the studies of bilinguals’ executive controls,
it is prudent to believe that bilinguals who code-switch frequently and
regularly, exercise the cognitive mechanisms underlying selective atten-
tion, such as attention to relevant aspects of a problem, inhibition of
attention to misleading information, and switching between competing
alternatives. Brain imaging studies also hinted at the potential relation
between code-switching and selective attention. For example, Jackson,
Swainson, Cunnington and Jackson (2001) found that similar inhibitory
mechanisms are involved in both response suppression and language
switching. One of the findings of a more recent study (Abutalebi et al.
2007) demonstrated that a switch into a less-exposed language triggered
activity in subcortical structures and the anterior cingulate cortex, which
are presumably involved in selective attention.
Moreover, if – as in the previous discussion – cognitive flexibility builds on
mechanisms of selective attention, one can expect that extensive code-
switching practice also contributes to flexibility in thinking. Lambert
(1977) suggested that bilingualism often entails repeated switching from
one language to another and constant dealing with several code systems
(phonological, grammatical, and lexical). Due to this experience, bilinguals
may learn to encode and access knowledge in diverse ways. This may
account for bilinguals’ greater metalinguistic awareness (e.g. Ben-Zeev

1977; Bialystok 1988; Cummins 1978) – the conscious awareness of struc-

tural properties of one’s language, the awareness of language itself, inde-
pendent of the message it is conveying (see Reynolds 1991). A positive effect
of bilingualism on metalinguistic abilities appears in understanding the
word–referent distinction (Ianco-Worrall 1972) and the arbitrariness of
language. People speaking several languages learn that the same concept
can have multiple referents in these languages. Having multiple labels will
force early separation of a word from its referent. As a result, multilingual
users have to conceptualize things according to their general properties
rather than relying on their linguistic symbols. This may lead, among other
things, to greater flexibility in thinking.
If having multiple linguistic referents for the same concept may boost
cognitive flexibility, experience with multiple sociocultural referents
may increase tolerance of ambiguity. Due to multicultural experience,
multilinguals can see the same phenomenon in several different ways
and have several perspectives on the same situation. Because different
cultural commonalities may provide different perspectives on the same
phenomena (Ricciardelli 1992b), these individuals may have greater tol-
erance of ambiguity. In addition, Lubart (1999) claimed that multilin-
guals’ advantages in divergent thinking may arise from the routine
ambiguity inherent in their multilingual practice, in which the same
basic idea may have different nuances in different languages. Thus,
cross-linguistic and cross-cultural experiences may advance multilin-
guals’ tolerance of ambiguity. Empirical data provide support for this
idea. Dewaele and Li Wei (2013) found that a high level of global profi-
ciency in various languages was linked to higher tolerance of ambiguity
scores. They also revealed that a stay abroad of more than three months
was linked to higher tolerance of ambiguity (although the effect leveled
off after one year). A similar pattern emerged from a study of Australian
students who had studied abroad (Bakalis and Joiner 2004). They obtained
higher tolerance of ambiguity scores than their counterparts who stayed
in their home institution.
This study also showed that the study abroad group achieved signifi-
cantly higher scores on openness, which indicates that an individual’s
multicultural experience may contribute to openness to new ideas and
experiences. This idea was supported by Leung and Chiu (2008), who
reported an interactive effect of multicultural experience and openness
to new experience on one’s creativity. Specifically, they found that hav-
ing extensive multicultural experience predicted better divergent think-
ing performance only among participants who were open to new
experience. Moreover, by analogy with Dewaele and Stavans’ (2014)
study reporting a correlation between proficiency in various languages
and open-mindedness, one can expect that mastery in multiple lan-
guages may be another contributor to an individual’s openness to
experience.

20.6 Fostering multi-competence and creative

competence in a classroom
Thus, the data reviewed present a case for a relationship between

multi-competence and creative competence. These competences can be
nurtured through education, and therefore discourse on multilingual
creativity would not be complete if it were not within the context of a
widely discussed topic in both multilingualism and creativity research that
arises from pedagogical considerations. It is evident that programs foster-
ing creativity operate separately from those offering multilingual instruc-
tion, and researchers and teachers have mutually exclusive training, being
educated in either creativity or language related disciplines but not in
both. Overall, the academic community generally disregards the potential
relationship between multilingualism and creativity (Kharkhurin 2012).
Similarly, the benefits of merging programs that foster creative potential
and multilingual abilities seem to escape the attention of educators.
However, the efficacy of a program combining both efforts can be directly
inferred from the research reviewed in this chapter. Various aspects of
individuals’ multi-competence have been demonstrated to facilitate cer-
tain cognitive mechanisms underlying their creative capacities. Therefore,
by combining multilingual and creative educational strategies, a far
greater synergy could be generated: a multilingual creative education
program would capitalize on the assets of both forms of education to
establish an effective and comprehensive curriculum. The need for this
type of program turns out to be immense, considering the outcomes of
scientific investigation, initiatives advanced by governmental policies,
and public opinion (see the review in Kharkhurin 2012).
Kharkhurin (2012) has proposed the Bilingual Creative Education (BCE)
program, which constitutes a unified teaching model introducing both
language learning and creativity-fostering instruction to the school curri-
culum. The rationale is not to establish a special program focusing on
children with exceptional abilities, but to suggest modifications to exist-
ing curricula and/or the classroom environment to promote bilingualism
and creativity in early schooling. The BCE program is grounded on several
conceptual premises. First, it disqualifies the elitist view and provides
opportunities to enhance the linguistic and creative capacities of all stu-
dents regardless of their intellectual and creative predispositions. This
entails the second characteristic of the program, its scope of application:
the BCE can be implemented in any school curriculum, depending on the
specific details of a given school. The role of the program coordinator
would be to modify the core of the program to reflect the specificity of
the student body and the economic, sociocultural, and political environ-
ment of each particular school. Instead of establishing a new school or a
special classroom with an entirely new curriculum, this program suggests

necessary modifications to convert any curriculum into one fostering

multi-competence and creative potential. Therefore, it reflects the EU
recommendation to its Member States (Marsh and Hill 2009) that meth-
odologies should be developed to modify and improve the effectiveness of
existing educational programs. Moreover, these modifications can be
accomplished at low cost because they would not require major restructur-
ing of existing school curricula. Third, the goal of the BCE program is to
facilitate multi-competence in a diversity of student populations. This
program is designed not only for migrants who speak their home language
and who are attempting to acquire the language of the country to which
they have migrated, but it is conceived for all children, immigrants as well
as those who want to acquire an additional language. Fourth, another goal
of the program is to foster children’s creative potential. The focus of the
program is not on bigger-C eminent creative achievements, but on the
smaller-c creative capacities (see Kaufman and Beghetto 2009) that are
grounded in mundane cognitive functioning and can be applied to every-
day problem-solving. The outcomes of this program do not reflect the
ambitious aspirations of nurturing eminent individuals (although this
perspective should not be excluded). Rather, the program aims at facilitat-
ing the overall linguistic, intellectual, and creative competences of young
children, thereby meeting the recommendations of certain government
policies (e.g. Commission of European Communities 2008).
Thus, the purpose of the BCE program is to introduce students to a
bilingual school curriculum and to foster the four defining aspects of
creativity (see note 2). To accomplish this goal, the program utilizes a
holistic approach that combines cognitive, personal, and environmental
factors. This approach considers not only educational aspects directly
pertinent to the school curriculum, but also those reflecting a child’s
personality and extracurricular settings. Note that the presentation of
the BCE program in the current chapter intends to stimulate the creative
thinking in education professionals rather than to provide an explicit
step-by-step description of the program. The interested reader can consult
Kharkhurin (2012), which presents a detailed description as well as theo-
retical and empirical considerations underlying the program.
20.7 Conclusion
The specific structure of multilingual memory may facilitate language-

mediated concept activation, which may in turn ensure simultaneous
activation of often unrelated concepts. At the same time, multilingual
practice may encourage inhibition and facilitation mechanisms of selec-
tive attention. These mechanisms seem to play an important role in diver-
gent and convergent thinking, and thereby foster an individual’s creative
performance. In addition, there is evidence that creative personality traits

such as cognitive flexibility, tolerance of ambiguity, openness to new

experience, and motivation can be developed as a result of multilingual
practice. The proposed cognitive mechanisms and personality traits appear
to benefit from multi-competence aspects such as proficiency in the lan-
guages used by an individual, the age of acquisition of these languages, the
circumstances and extent to which an individual switches between these
languages, and the sociocultural environment and emotional context in
which these languages are acquired and used. It is evident that empirical
data suggesting links between multi-competence and creativity are highly
scattered and often speculative. Overall, multilingual creativity lacks sys-
tematic research, especially research focusing on variation in multilinguals’
creative capacities. Therefore, systematic investigation of the proposed and
possibly other factors in multilingual creativity is required.
At this point only one conclusion can be certain: the use of multiple
languages has important ramifications for the human mind as a whole. It
develops multiple competences, which facilitate cognitive abilities. In our
case, it is likely to improve an individual’s creative capacities. As was
mentioned in Section 20.1, creativity is prompted by a large variety of
factors such as education, expertise, motivation, attitudes, personality
traits, personal experience, and socioeconomic and sociocultural condi-
tions. Although the effect of multi-competence per se may be overridden by
those factors, it may play a substantial role in their development.
Therefore, the findings reviewed in this chapter should stimulate new
ideas and serve as a benchmark for extensive future research in multi-
lingual creativity and education.
Notes
1. This perspective rests on the creative cognition paradigm, which

assumes creative capacity to be an essential property of normative
human cognition (Ward, Smith and Finke 1999). Creative ideas ‘emerge
from the application of ordinary, fundamental cognitive processes to
existing knowledge structures’ (Ward 2007, p. 28). Beyond the obvious
examples of artistic, scientific, and technological advancement that are
usually listed as instances of creativity, there is the subtler, but equally
compelling generativity associated with everyday thought. One of the
most striking examples of generativity is the productivity of language:
we are able to construct an infinite number of grammatical sentences
using a limited number of words and a small set of rules (Chomsky
1972). The generated sentences can be unique both to an individual’s
linguistic practice and to the common practice of a given language. In
line with the premises of creative cognition, the generativity goes
beyond everyday human cognition and satisfies the criteria of creative
products: novelty and utility.

2. Kharkhurin (2014) challenged this definition as being biased by a Western

perception of creativity. He proposed an alternative four-criterion con-
struct of creativity, which in addition to novelty and utility considers two
other characteristics typical for an Eastern perception of creativity: aes-
thetics and authenticity.
References
Abutalebi, J., Brambati, S. M., Annoni, J.-M., Moro, A., Cappa, S. F. and
Perani, D. 2007. ‘The neural cost of the auditory perception of language
switches: an event-related functional magnetic resonance imaging study
in bilinguals’, The Journal of Neuroscience 27, 13762–13769.
Adi-Japha, E., Berberich-Artzi, J. and Libnawi, A. 2010. ‘Cognitive flexibility
in drawings of bilingual children’, Child Development 81, 1356–1366.
Altarriba, J. and Basnight-Brown, D. M. 2007. ‘Methodological considera-
tions in performing semantic and translation priming experiments
across languages’, Behavior Research Methods, Instruments, and Computers
39, 1–18.
Amabile, T. M. 1996. Creativity in Context: Update to ‘The Social Psychology of
Creativity’. Boulder, CO: Westview Press.
Androutsopoulos, J. 2015. ‘Networked multilingualism: some language
practices on Facebook and their implications’, International Journal of
Averill, J. R., Chon, K. K. and Hahn, D. W. 2001. ‘Emotions and creativity,
East and West’, Asian Journal of Social Psychology 4,165–183.
Bakalis, S. and Joiner, T. 2004. ‘Participation in tertiary SA programs: the
role of personality’, International Journal of Educational Management 18,
286–291.
Ben-Zeev, S. 1977. ‘The influence of bilingualism on cognitive strategy and
cognitive development’, Child Development 48, 1009–1018.
Bhatia, T. K. and Ritchie, W. C. 2008. ‘The bilingual mind and linguistic
creativity’, Journal of Creative Communications 3, 5–21.
Bialystok, E. 1988. ‘Levels of bilingualism and levels of linguistic aware-
ness’, Developmental Psychology 24, 560–567.
Bialystok, E. 2001. Bilingualism in Development: Language, Literacy, and
Bialystok, E. 2005. ‘Consequences of bilingualism for cognitive develop-
ment’, in J. F. Kroll and A. M. B. de Groot (eds), pp. 417–432.
Bialystok, E., Craik, F. I. M., Klein, R. and Viswanathan, M. 2004.
‘Bilingualism, aging, and cognitive control: evidence from the Simon
task’, Psychology and Aging 19, 290–303.
Bialystok, E. and Martin, M. M. 2004. ‘Attention and inhibition in bilingual
children: evidence from the dimensional change card sort task’,
Developmental Science 7, 325–339.

Bialystok, E. and Shapero, D. 2005. ‘Ambiguous benefits: the effect of

bilingualism on reversing ambiguous figures’, Developmental Science 8,
595–604.
Birman, D., Trickett, E. J. and Vinokurov, A. 2002. ‘Acculturation and
adaptation of Soviet Jewish refugee adolescents: predictors of adjust-
ment across life domains’, American Journal of Community Psychology 30,
585–607.
Budner, S. 1962. ‘Intolerance of ambiguity as a personality variable’, Journal
of Personality 30, 29–50.
Carlson, S. M. and Meltzoff, A. N. 2008. ‘Bilingual experience and executive
functioning in young children’, Developmental Science 11, 282–298.
Carringer, D. C. 1974. ‘Creative thinking abilities of Mexican youth: the
relationship of bilingualism’, Journal of Cross-Cultural Psychology 5,
492–504.
Chik, A. 2010. ‘Creative multilingualism in Hong Kong popular music’,
World Englishes 29, 508–522.
Chomsky, N. 1972. Language and Mind. New York: Harcourt Brace
Jovanovich.
Commission of European Communities 2008. ‘Improving competences for
the 21st century: an agenda for European cooperation on schools’, in
Communication from the Commission to the European Parliament, the Council, the
European Economic and Social Committee and the Committee of the Regions.
Brussels.
Costa, A., Hernandez, M. and Sebastián-Gallés, N. 2008. ‘Bilingualism aids
conflict resolution: evidence from the ANT task’, Cognition 106, 59–86.
Cropley, A. 2006. ‘In praise of convergent thinking’, Creativity Research
Journal 18, 391–404.
Cummins, J. 1976. ‘The influence of bilingualism on cognitive growth: a
synthesis of research findings and explanatory hypothesis’, Working
Papers on Bilingualism 9, 1–43.
Cummins, J. 1978. ‘Bilingualism and the development of metalinguistic
awareness’, Journal of Cross-Cultural Psychology 9, 131–149.
Cummins, J. and Gulutsan, M. 1974. ‘Some effects of bilingualism on
cognitive functioning’, in S. T. Carey (ed.), Bilingualism, Biculturalism and
Education: Proceedings from the Conference at Collège Universitaire Saint-Jean,
pp. 129–136. Edmonton: University of Alberta.
Cushen, P. J. and Wiley, J. 2011. ‘Aha! Voila! Eureka! Bilingualism and
insightful problem solving’, Learning and Individual Differences 21,
458–462.
Dehaene, S., Dupoux, E., Mehler, J., Cohen, L., Paulesu, E., Perani, D., van de
Moortele, P.-F., Lehericy, S. and Le Bihan, D. 1997. ‘Anatomical variabil-
ity in the cortical representation of first and second language’,
Neuroreport: An International Journal for the Rapid Communication of Research
in Neuroscience 8, 3809–3815.

Dewaele, J.-M. 2010. Emotions in Multiple Languages. Basingstoke: Palgrave

Macmillan.
tolerance of ambiguity?’, Bilingualism: Language and Cognition 16,
231–240.
guals’, International Journal of Bilingualism 18 (3), 203–221.
Dewaele, J.-M. and van Oudenhoven, J. P. 2009. ‘The effect of multilingual-
ism/multiculturalism on personality: no gain without pain for Third
Culture kids?’, International Journal of Multilingualism 6, 443–459.
Duncker, K. 1945. ‘The structure and dynamics of problem-solving pro-
cesses’, Psychological Monographs 58, 1–112.
Engin, A. O. 2009. ‘Second language learning success and motivation’,
Social Behavior and Personality 37, 1035–1042.
European Commission 2012. ‘Europeans and their languages’,
Eurobarometer Special, 386. http://ec.europa.eu/public_opinion/archives/
ebs/ebs_386_en.pdf
Feist, G. J. 1998. ‘A meta-analysis of personality in scientific and artistic
creativity’, Personality and Social Psychology Review 2, 290–309.
Florida, R. L. 2002. The Rise of the Creative Class: and How it’s Transforming Work,
Leisure, Community and Everyday Life. New York: Basic Books.
Florida, R. L., Mellander, C. and Stolarick, K. 2011. Creativity and Prosperity:
The Global Creativity Index. Toronto, Canada: Martin Prosperity Institute.
Education. Basingstoke: Palgrave.
Genesee, F. 1978. ‘Language processing in bilinguals’, Brain and Language 5,
1–12.
Gordon, M. M. 1964. Assimilation in American Life: The Role of Race, Religion, and
National Origins. Oxford University Press.
Green, D. W. 1998. ‘Mental control of the bilingual lexico-semantic sys-
tem’, Bilingualism: Language and Cognition 1, 67–81.
Guilford, J. P. 1967. The Nature of Human Intelligence. New York: McGraw-Hill.
Guilford, J. P. 1968. Intelligence, Creativity, and their Educational Implications.
San Diego, CA: R. R. Knapp.
Hommel, B., Colzato, L. S., Fischer, R. and Christoffels, I. 2011.
‘Bilingualism and creativity: benefits in convergent thinking come
with losses in divergent thinking’, Frontiers in Psychology 2, arti-
cle 273.
Ianco-Worrall, A. D. 1972. ‘Bilingualism and cognitive development’, Child
Jackson, G. M., Swainson, R., Cunnington, R. and Jackson, S. R. 2001. ‘ERP
correlates of executive control during repeated language switching’,

Jacobs, J. F. and Pierce, M. L. 1966. ‘Bilingualism and creativity’, Elementary

English 40, 499–503.
Jonsson, C. 2005. Code-switching in Chicano Theater. Umeaº , Sweden: Umeaº
University Press.
Karapetsas, A. and Andreou, G. 1999. ‘Cognitive development of fluent and
nonfluent bilingual speakers assessed with tachistoscopic techniques’,
Psychological Reports 84, 697–700.
Kasof, J. 1997. ‘Creativity and breadth of attention’, Creativity Research
Journal 10, 303–315.
Kaufman, J. C. and Beghetto, R. A. 2009. ‘Beyond big and little: the four c
model of creativity’, Review of General Psychology 13, 1–12.
Kaufmann, G. 2003. ‘Expanding the mood-creativity equation’, Creativity
Research Journal 15, 131–135.
Kharkhurin, A. V. 2007. ‘The role of cross-linguistic and cross-cultural
experiences in bilinguals’ divergent thinking’, in I. Kecskes and L.
Albertazzi (eds), Cognitive Aspects of Bilingualism, pp. 175–210. Dordrecht,
Netherlands: Springer.
Kharkhurin, A. V. 2008. The effect of linguistic proficiency, age of second
language acquisition, and length of exposure to a new cultural environ-
ment on bilinguals’ divergent thinking’, Bilingualism: Language and
Kharkhurin, A. V. 2009. ‘The role of bilingualism in creative performance
on divergent thinking and Invented Alien Creatures tests’, Journal of
Creative Behavior 43, 59–71.
Kharkhurin, A. V. 2010. ‘Sociocultural differences in the relationship
between bilingualism and creative potential’, Journal of Cross-Cultural
Kharkhurin, A. V. 2011. ‘The role of selective attention in bilingual crea-
tivity’, Creativity Research Journal 23, 239–254.
Kharkhurin, A. V. 2012. Multilingualism and Creativity. Bristol, UK:
Kharkhurin, A. V. 2014. ‘Creativity.4in1: four-criterion construct of crea-
tivity’, Creativity Research Journal 26, 338–352.
Kharkhurin, A. V. and Altarriba, J., 2015. ‘The effect of mood on bilingual
creativity’, Bilingualism: Language and Cognition. Available on CJO 2015 doi:
10.1017/51366728915000528.
Kharkhurin, A. V. and Isaeva, M. G. 2015. Language mediated concept
activation in bilingual memory as a facilitation mechanism of creative
behavior. Paper presented at the The 10th International Symposium on
Bilingualism, Rutgers University, NJ.
Kharkhurin, A. V. and Li Wei 2014. ‘The role of codeswitching in bilingual
creativity’, International Journal of Bilingual Education and Bilingualism.
doi: 10.1080/13670050.2014.88211.

Kharkhurin, A. V. and Samadpour Motalleebi, S. N. 2008. ‘The impact of

culture on the creative potential of American, Russian, and Iranian
college students’, Creativity Research Journal 20, 404–411.
Kim, K. H. S., Relkin, N. R., Lee, K.-M. and Hirsch, J. 1997. ‘Distinct cortical
Koestler, A. 1964. The Act of Creation. Lewiston, NY: Macmillan.
Konaka, K. 1997. The Relationship Between Degree of Bilingualism and Gender to
Divergent Thinking Ability Among Native Japanese-speaking Children in the New
York Area. Dissertation, New York University. Available from Univ
Microfilms International database.
Kostandyan, М. E. and Ledovaya, Y. A. 2013. ‘How the age of language
acquisition relates to creativity’, Procedia: Social and Behavioral Sciences
86, 140–145.
Kroll, J. F. and de Groot, A. M. B. 1997. ‘Lexical and conceptual memory in
the bilingual: mapping form to meaning in two languages’, in A. M. B. de
Groot and J. F. Kroll (eds), Tutorials in Bilingualism: Psycholinguistic
Perspectives, pp. 169–199. Hillsdale, NJ: Erlbaum.
Psycholinguistic Approaches. Oxford University Press.
Lamb, S. 1999. Pathways of the Brain. Amsterdam: John Benjamins.
Lambert, W. E. 1956. ‘Developmental aspects of second-language acquisi-
tion, part I: associational fluency, stimulus provocativeness, and word-
order influence’, The Journal of Social Psychology 43, 83–89.
Lambert, W. E. 1977. ‘The effect of bilingualism on the individual: cogni-
tive and social consequences’, in A. Hornby (ed.), Bilingualism:
Psychological, Social and Educational Implications, pp. 15–27. New York:
Academic Press.
Lee, H. and Kim, K. H. 2011. ‘Can speaking more languages enhance your
creativity? Relationship between bilingualism and creative potential
among Korean American students with multicultural link’, Personality
and Individual Differences 50, 1186–1190.
Lemmon, C. R. and Goggin, J. P. 1989. ‘The measurement of bilingualism and
its relationship to cognitive ability’, Applied Psycholinguistics 10, 133–155.
Leung, A. K.-Y. and Chiu, C.-Y. 2008. ‘Interactive effects of multicultural
experiences and openness to experience on creative potential’, Creativity
Research Journal 20, 376–382.
Leung, A. K.-Y., Maddux, W. W., Galinsky, A. D. and Chiu, C.-Y. 2008.
‘Multicultural experience enhances creativity: the when and how’,
American Psychologist 63, 169–181.
Li Wei and Wu, C.-J. 2009. ‘Polite Chinese children revisited: creativity and
the use of codeswitching in the Chinese complementary school class-
room’, International Journal of Bilingual Education and Bilingualism 12,
193–211.
Lubart, T. I. 2000. ‘Models of the creative process: past, present and future’,
Creativity Research Journal 13, 295–308.

Maddux, W. W. and Galinsky, A. D. 2009. ‘Cultural borders and mental

barriers: the relationship between living abroad and creativity’, Journal of
Personality and Social Psychology 96, 1047–1061.
Marsh, D. and Hill, R. 2009. Study on the Contribution of Multilingualism to
Creativity: Final Report. Brussels: European Commission.
Masgoret, A. M. and Gardner, R. C. 2003. ‘Attitudes, motivation, and second
language learning: a meta-analysis of studies conducted by Gardner and
associates’, Language Learning 53, 167–210.
McClelland, J. L. and Rumelhart, D. E. 1985. ‘Distributed memory and the
representation of general and specific information’, Journal of
Experimental Psychology: General 114, 159–188.
McCrae, R. R. and Costa Jr, P. T. 1997. ‘Conceptions and correlates of
openness to experience’, in R. Hogan, J. Johnson and S. Briggs (eds),
Handbook of Personality Psychology, pp. 825–847. San Diego: Academic
Press.
McLaughlin, B. 1984. Second-language Acquisition in Childhood, vol. 1: Preschool
Children, second edition. Hillsdale, NJ: Erlbaum.
Mednick, S. A. 1962. ‘The associative basis of the creative process’,
Psychological Review 69: 220–232.
Mendelsohn, G. A. 1976. ‘Associative and attentional processes in creative
performance’, Journal of Personality 44, 341–369.
Meyer, D. E. and Schvaneveldt, R. W. 1971. ‘Facilitation in recognizing
pairs of words: evidence of a dependence between retrieval operations’,
Journal of Experimental Psychology 90, 227–234.
Montgomery, D., Hodges, P. A. and Kaufman, J. S. 2004. ‘An exploratory
study of the relationship between mood states and creativity self-percep-
tions’, Creativity Research Journal 16, 341–344.
Niu, W. and Sternberg, R. J. 2001. ‘Cultural influences on artistic creativity
and its evaluation’, International Journal of Psychology 36, 225–241.
Okoh, N. 1980. ‘Bilingualism and divergent thinking among Nigerian and
Welsh school children’, Journal of Social Psychology 110, 163–170.
Oxford, R. L. and Ehrman, M. 1992. ‘Second language research on indivi-
dual differences’, Annual Review of Applied Linguistics 13, 188–205.
Paradis, M. 1997. ‘The cognitive neuropsychology of bilingualism’, in
A. M. B. de Groot and J. F. Kroll (eds), Tutorials in Bilingualism:
Psycholinguistic Perspectives, pp. 331–354. Hillsdale, NJ: Erlbaum.
Pavlenko, A. 2000. ‘New approaches to concepts in bilingual memory’,
Pavlenko, A. (ed.) 2006. Bilingual Minds: Emotional Experience, Expression, and
Representation. Bristol, UK: Multilingual Matters.
Reynolds, A. G. 1991. Bilingualism, Multiculturalism, and Second Language
Learning: The McGill Conference in Honour of Wallace E. Lambert. Hillsdale,
NJ: Erlbaum.

Ricciardelli, L. A. 1992a. ‘Bilingualism and cognitive development in

relation to threshold theory’, Journal of Psycholinguistic Research 21,
301–316.
Ricciardelli, L. A. 1992b. ‘Creativity and bilingualism’, Journal of Creative
Behavior 26, 242–254.
Rubin, J. 1975. ‘What the “good language learner” can teach us’, TESOL
Quarterly 9, 41–51.
Runco, M. A. 1991. Divergent Thinking. Westport, CT: Ablex.
Runco, M. A. 2014. Creativity: Theories and Themes: Research, Development, and
Practice, second edition. Boston, MA: Elsevier Academic Press.
Runco, M. A., Dow, G. and Smith, W. R. 2006. ‘Information, experience, and
divergent thinking: an empirical test’, Creativity Research Journal 18,
269–277.
Russ, S. W. 1999. Affect, Creative Experience, and Psychological Adjustment.
Philadelphia: Brunner, Mazel.
Sebba, M., Jonsson, C. and Mahootian, S. 2012. Language Mixing and Code-
switching in Writing : Approaches to Mixed-language Written Discourse. New
York: Routledge.
Silvia, P. J., Nusbaum, E. C., Berg, C., Martin, C. and O’Connor, A. 2009.
‘Openness to experience, plasticity, and creativity: exploring lower-
order, high-order, and interactive effects’, Journal of Research in
Personality 43, 1087–1090.
Srivastava, A. L. and Khatoon, R. 1980. ‘Effect of difference between
mother tongue and another tongue as medium of instruction on
achievement, mental ability and creativity of the Vll standard children’,
in E. Annamalai (ed.), Bilingualism and Achievement in Schools. Mysore: CIIL.
Sternberg, R. J. (ed.) 1999. Handbook of Creativity. Cambridge University
Press.
Torrance, E. P., Gowan, J. C., Wu, J.-J. and Aliotti, N. C. 1970. ‘Creative
functioning of monolingual and bilingual children in Singapore’,
Journal of Educational Psychology 61, 72–75.
Torrance, E. P. and Safter, H. T. 1999. Making the Creative Leap Beyond.
Buffalo, NY: Creative Education Foundation Press.
Wang, F. X. 2008. ‘Motivation and English achievement: an exploratory
and confirmatory factor analysis of a new measure for Chinese stu-
dents of English learning’, North American Journal of Psychology 10,
633–646.
Ward, T. B. 1994. ‘Structured imagination: the role of category structure in
exemplar generation’, Cognitive Psychology 27, 1–40.
Ward, T. B. 2007. ‘Creative cognition as a window on creativity’, Methods 42,
28–37.
Ward, T. B., Smith, S. M. and Vaid, J. 1997. Creative Thought: An Investigation of
Conceptual Structures and Processes. Washington, DC: American
Psychological Association.

Zenasni, F., Besançon, M. and Lubart, T. I. 2008. ‘Creativity and tolerance of

ambiguity: an empirical study’, Journal of Creative Behavior 42, 61–73.
Zha, P., Walczyk, J. J., Griffith-Ross, D. A., Tobacyk, J. J. and Walczyk, D. F.
2006. ‘The impact of culture and individualism–collectivism on the
creative potential and achievement of American and Chinese adults’,
Creativity Research Journal 18, 355–366.

21
language teaching
Virginia M. Scott
A second language is not just adding rooms to your house by

building an extension at the back: it is rebuilding all the internal
walls. (Vivian Cook 2005)
In his 2005 presentation at the Second Language Research Forum at

Columbia University in New York, Vivian Cook used this clever image to
help us envision the mind as a house with more than one language under
one roof. The house-as-mind evokes an ecosystem in which languages live
in mutual interdependence; the doors in the house connect all the rooms
so that movement is effortless and uninhibited. Above all, it offers a
picture of all the languages a person knows and uses living in the same
house, with none consigned to the extension at the back. This image
captures Cook’s notion of the multi-competent mind that accommodates
the knowledge and use of more than one language.
This architectural metaphor has dwelled in my imagination for nearly a
decade and the construction project is far from complete! I am still grap-
pling with questions related to how the multi-competence framework offers
us different ways of thinking about classroom foreign language learning. In
particular, I have tried to envision the classroom as part of the main house
and not as the “extension at the back.” All too often real language learning
(or acquisition) is believed to occur in natural, immersion settings outside
the classroom; the classroom, by contrast, is largely considered an artificial
setting for language learning. At best, it may be viewed as the staging area
for real language use in the natural setting of the target culture. What if we
didn’t privilege the natural, immersion setting over the formal, instructed
setting? What if the classroom were the house and not the extension?
In this chapter I argue that the foreign language classroom is a privileged
environment in which extraordinary learning experiences can transpire.
Rather than concentrate with single-minded attention on the develop-
ment of monolingual target language proficiency, I propose that the

446 VI RGI N IA M. SCOT T
classroom is the ideal setting for teachers and students to address the
broader implications of language learning. In order to make this case, I
use Cook’s definition of the multi-competent L2 user to theorize a multi-
competent L2 learner. I then suggest the ways the multi-competence
perspective can be put into practice in the foreign language classroom.
Because I consider intentional, guided reflection to be at the heart of what
it means to be a multi-competent L2 learner, I explain what I mean by
“reflection” and “awareness” in this context. Finally, I review a study
designed to assess the effects of awareness on students’ sense of
themselves as multi-competent L2 users. I conclude with an argument
that the multi-competence perspective is central to ethical and principled
language education, and ultimately to social justice in the world.
21.1 Defining the multi-competent L2 learner
Multi-competence concerns the mind of any user of a second

language at any level of achievement.
(Vivian Cook, Chapter 1, this volume, p. 5)
Multi-competence, as defined by Cook in this volume and elsewhere, is the

overall system of a mind that knows and uses more than one language.
Rather than a theory or model of second language acquisition, it is a per-
spective, or a way of looking at acquisition and use of multiple languages.
“Multi-competence is a way of looking at things from another angle rather
than of exploring the implications and contradictions within the same
perspective” (Cook, Chapter 1, this volume, p. 3). This perspective main-
tains, above all, that people who know more than one language have a
distinct compound state of mind that is not the equivalent of two (or
more) monolingual states. It is a holistic view of language development
and use that eschews bilingualism or multilingualism as some idealized
condition characterized by advanced language proficiency in more than
two languages. Multi-competence is a dynamic system that accounts for
the natural ebb and flow of a person’s native language as well as
other languages in various stages of development (Scott 2010, p. 17).
Moreover, “Multi-competence affects the whole mind, i.e. all language and
cognitive systems, rather than language alone” (Cook, Chapter 1, this
volume, p. 15).
The most important contribution of the multi-competence perspective
is the notion of the L2 user, defined as “any person who uses another
language than his or her first language (L1), that is to say, the one learnt
first as a child” (Cook 2002b, p. 1). The L2 user is, therefore, a unique,
individual speaker–hearer of a second language who stands in contrast to a
native speaker of that language. It is important to note that although the
term “native speaker” is part of common parlance, many current

Multi-competence and language teaching 447
practitioners consider it an inadequate and perhaps even unacceptable

description. As Kramsch (1998) noted:
The “native speaker” of linguists and language teachers is in fact an
abstraction based on arbitrarily selected features of appearance and
demeanor . . . The native speaker is, moreover, a monolingual, monocul-
tural abstraction: he/she is one who speaks only his/her (standardized)
native tongue and lives by one (standardized) national culture. In reality,
most people partake of various languages or language varieties and live by
various cultures and subcultures. (Kramsch 1998, pp. 79–80)
Given this understanding, multi-competent L2 use, rather than native

speaker-ness, is the standard. Idiosyncratic second language use at any
stage of development by a particular individual in a specific context takes
precedence over idealized native language use; the local L2 user supplants
the universal native speaker.
Multi-competent L2 use refers to both the total knowledge of languages in
one mind and the use of a second language that is not an imitation of one’s
first language but rather operating on its own terms alongside the first
language (Cook 1995). These multi-competent L2 users can be found every-
where in cities across the world. In my own region I interact regularly with a
variety of multi-competent L2 users who speak southern American English:
my friend’s husband, an Ethiopian taxi driver; my Korean–American daugh-
ter-in-law’s mother, owner of a laundry facility; a Haitian medical practi-
tioner at our university hospital; my Parisian colleague, a professor of
French feminist thought; my Chinese graduate student studying second
language acquisition and foreign language pedagogy in preparation for his
return to China to teach English. All of these people have marked non-native
accents and make repeated, systematic errors when they speak English; all
of them are successful and productive members of our community who
speak, read, and write English with varying degrees of accuracy.
Multi-competence is, by and large, a term that refers to people who are
using their languages in real-life settings in order to get along. They are
very likely motivated by both the need and the desire to communicate
with the people they encounter in their daily lives. Much of this language
use is likely to be spontaneous and produced in response to immediate,
real-life stimuli. Although there may be some conscious reflection about
what it means to know and use several languages, I would argue that multi-
competent L2 users in immersion settings are mostly unaware of the ways
their languages operate in tandem. Even if they are studying in a classroom
in order to improve their proficiency in the target language – for example
English, in the case of my friends mentioned above – they are likely to be
oblivious to their multi-competence. They have no reason to think about it.
Unlike people in an immersion setting, the foreign language learner in a
classroom may have fifty minutes three times a week to hear and practice
the target language. Although learners come to the language classroom

with a wide variety of goals (i.e., completing a school requirement, getting

a good grade, wanting to travel, hoping for a good job), the principal goal of
their work in the classroom is not survival. They (and their teachers) are
generally focused on developing monolingual, target language profi-
ciency. As I stated in Double Talk, “Achieving some level of bilingual func-
tioning is clearly a valid goal of foreign language instruction; however, it is
ironic that becoming bilingual often means engaging primarily in a mono-
lingual encounter with the target language” (Scott 2010, p. 3). If we agree
that multi-competence is a compound state of mind that is not the equiva-
lent of two (or more) monolingual states, and the foreign language class-
room generally privileges a monolingual approach, I would argue that
foreign language learners in a classroom do not spontaneously become
multi-competent L2 users. Instead, many of them abandon their studies,
doubting their abilities to learn a foreign language and feeling like perpe-
tually deficient native-speakers.
Although Cook rejects the term “L2 learner” because he thinks it
relegates people to a subordinate status of learners as opposed to L2 users
(see Cook, Chapter 1, this volume), I regard the classroom as an ideal setting
for exploring the essential connections between the L2 learner and the L2
user. Understanding that a successful L2 user is not a deficient native
speaker can be enormously empowering to a foreign language learner.
Likewise, exploring the theoretical construct of the multi-competent L2
user can inspire foreign language learners to continue their studies, even
when faced with the inevitable challenges of language learning. In my
work (Scott 2010) I use the term “multi-competent second language lear-
ner” to describe learners who reflect critically about language learning and
language use. Rather than thinking of native-like speech as the central goal
(an unattainable goal for many learners), the multi-competent L2 learner
understands what it means to be an L2 user and is using the classroom to
develop an understanding of who they are in our multilingual, multicul-
tural world. Ultimately, I would argue that teachers should incorporate an
intentional focus on developing multi-competent L2 learners so that the
classroom becomes the place where multi-competent L2 use emerges
organically.
21.2 Multi-competence and the classroom
As the multi-competence approach developed and broadened,

it became evident that it was more a perspective from which to
view the acquisition and use of multiple languages than a
theory or a model. (Vivian Cook, Chapter 1, this volume, p. 3)
This volume is devoted to defining and assessing the many and varied
implications of multi-competence, and the contributors would all likely

agree that multi-competence is a notion that has value and contributes to

a rethinking of language acquisition and language use. However, multi-
competence as a “perspective from which to view the acquisition and use
of multiple languages,” as Cook states in the quote above, is not a concept
that has informed foreign language teaching practice. In general, current
practice adheres to a chiefly monolingual encounter with the target lan-
guage and students are encouraged to focus on developing proficiency in the
target language. Moreover, this monolingual approach is often designed to
promote native-like target language proficiency: “the goal set for students
was mostly to get as close as possible to monolingual native speakers” (Cook
2002c, p. 329). If we consider multi-competence to be a valid, natural, and
even desirable state of being, then both the approach and the learning goals
for classroom foreign language learning must be re-envisioned.
There is no question that classroom foreign language learning should
focus on helping students learn the target language. However, there is little
consensus about what actually constitutes success. That is, what does it
mean to be a successful language learner? Success may be measured by
achievement in a particular course of study or scores on a standardized test;
it may be assessed by a student’s level of proficiency (novice, intermediate,
advanced, superior, distinguished) in particular skills according to the
ACTFL scale. Macaro (2010b) challenges us to think carefully about what
constitutes language learning success and how we can measure it. He argues
that research has yet to fully explore the notion of the “good language
learner” and the criteria used to characterize good language learning:
An understanding of the meaning of language learning success, just like
an understanding of what it is to really know a word, is fundamental to
language teaching pedagogy. For a teacher to have a clear idea of what the
end and intermediate goals of learning are is crucial to the relationship
s/he builds up with his/her students and to the possibility of adapting
his/her language curriculum and pedagogy. (Macaro 2010b, p. 304)
Although there are no simple answers to questions regarding what we

should be teaching in the foreign language classroom and how we should
measure success, I believe the multi-competence perspective offers us a
framework for this rethinking.
Assuming a multi-competence perspective in the foreign language class-
room involves a careful reconsideration of three key issues. First, teachers
must adopt a multilingual stance rather than a monolingual stance. That
is, teachers should see learners as individuals who are already members of
a community that uses more than one language rather than as monolin-
guals who are adding a new language to their first language. A classroom
that uses multi-competence as its framework treats “the diverse languages
of the community as a coherent whole rather than separately” (Cook,
Chapter 1, this volume, p. 3). In practical terms, this principle involves
listening to the language stories of the students in a classroom, creating

a community of learners who know about one another’s language his-

tories – the languages spoken by parents or grandparents, incidental
encounters with speakers of other languages, feelings about speakers of
particular languages, and so on. Acknowledging that the target language is
not the only language present in the classroom is essential and can
sensitize students to the notion of multi-competence. This multilingual
stance is not a one-time community-building activity; rather it should be
cultivated throughout a course of study, with each new topic (linguistic or
cultural) being considered through this lens. At times these discussions
may be in the shared language of the students (English in the US class-
room), whereas at other times it may be possible to use the target language.
Regardless, this multilingual stance endorses the legitimacy of all
languages – those spoken by people in power as well as those spoken by
marginalized people – thereby creating a community of learners who learn
to respect L2 use in its many forms.
It is worth mentioning that foreign language teachers are often reluc-
tant to use the L1 in the classroom for fear that students will fall back on
the L1 and circumvent the L2. Cook (2002c) has argued that we should
adopt a bilingual approach to teaching, arguing that teachers “should
develop the systematic use of the L1 in the classroom alongside the L2 as
a reflection of the realities of the classroom situation, as an aid to learning
and as a model for the world outside” (p. 332). This approach may seem
counterproductive to many teachers, especially given general consensus
about the importance of input and interaction in the target language.
However, it is Cook’s use of the word “systematic” that must be considered
carefully. Random and arbitrary use of the L1 should be discouraged. For
example, a teacher who consistently translates what s/he says or offers
regularly to be a “walking dictionary” does not help students use their
intuition and problem-solving skills to decipher and interpret the target
language. Rather, teachers should bear in mind the ways that L1 and L2 are
used in real life by people who speak the same languages; bilingual people
engage in word play, they code-switch, and they search for novel and
creative ways to communicate a message. The L1 should be acknowledged
and present in the foreign language classroom, but teachers and students
alike should have a conscious understanding of why and how it is to be
used.
The second issue for consideration with regard to multi-competence
perspective in the classroom involves Cook’s notion of the L2 user, defined
as a person who uses a language other than his/her first. An L2 user is not
the equivalent of two monolingual native speakers; rather s/he is a unique
person who has a distinct set of language experiences that come into play
when s/he uses the second language. Classrooms are full of people who are
L2 users in their daily lives yet the formal learning environment typically
rejects the value of L2 user-ness and reveres the idealized native speaker.
Appreciating the L2 user would require that classroom teaching not focus

on native speakers of one standardized version of a particular language,

such as Parisian French or British English.
Students need to be shown the richness of L2 use. Rather than a few L2
users stumbling through conversations with powerful native speakers,
they need to encounter the language of people who use the language
effectively as a second language, who, because they speak two languages,
can say things that monolingual native speakers can never say.
(Cook 2002c, p. 338)
It should be noted that many foreign language teachers are L2 users – that
is, not native speakers of the language(s) they teach – and may have
distinctive accents and possibly unique writing styles. L2 user-ness is
not a shameful condition but rather normal, natural, and even creative.
With regard to classroom practice, accepting L2 use means that pronun-
ciation and linguistic accuracy cannot be the principal measures of
success. Teachers must emphasize the many kinds of L2 use: listening
to authentic texts of all kinds, reading for information and for pleasure,
browsing the internet, chatting informally using social media, and so on.
Ultimately, as Macaro states,
As language teachers we should, of course, be trying to create bilinguals

and not native speakers of the L2. This requires a fundamental shift in how
teachers need to conceptualise the L2 learning curriculum and pedagogy.
(Macaro 2010b, p. 301)
Awareness is the third and most important issue when adopting a multi-
competence perspective in the foreign language classroom. Learners in a
classroom do not spontaneously become multi-competent L2 users; in fact, a
monolingual target-language classroom environment can inadvertently
impede the development of multi-competence. Multi-competence is a state
of mind and foreign language teachers need to cultivate an awareness of
that state of mind, of that perspective, of that stance, for themselves and for
their students. In my view, we need to train people to think differently about
language learning and language use; awareness, or a conscious understand-
ing of the notion of multi-competence, is at the core of this approach.
21.3 Language awareness and the multi-competent

L2 learner
[T]he task for the language educator is above all to educate, to

promote an ability to change perspective and to challenge what
is taken for granted. (Byram 2008, p. 17)
Fundamental to the multi-competence perspective in the foreign language

classroom is the idea that teachers have a responsibility to do more
than teach the target language. As indicated previously, developing

multi-competence in the classroom involves adopting a multilingual

stance, validating the L2 user, and helping students understand who they
can be as multi-competent L2 users. Although teachers might argue that
these tasks are beyond the scope of foreign language teaching, I would
argue that we are ideally suited for this work precisely because we are
language specialists who care about how our students conceive of them-
selves as users of their newly learned second languages. Byram’s (2008,
2012) distinction between language learning and language education clari-
fies the point. He contends that language learning involves the develop-
ment of proficiency in a particular language whereas language education
has social and political purposes. Language education goes beyond the
scope of language learning and seeks to raise students’ awareness about
what it means to be a member of a linguistically and culturally diverse
world. To achieve the goals of language education, Byram proposes a
model that “represents language and culture competence holistically and
shows the relationship between language competence – including lan-
guage awareness – and intercultural competence, including cultural
awareness” (Byram 2012, p. 6). This distinction between language learning
and language education challenges foreign language teachers to take a
broader view of their role and to address both dimensions.
In my view, language education, with its focus on fostering language
awareness, is an essential part of a multi-competence perspective in the
foreign language classroom. Although there are several definitions of
language awareness, including a person’s declarative knowledge of
underlying systems of language (i.e. grammar), I am using the term “lan-
guage awareness” to refer to “a person’s sensitivity to and conscious
awareness of the nature of language and its role in human life” (Svalberg
2007, p. 288). To understand multi-competence and the notion of a multi-
competent L2 user, students must be guided to reflect critically about
themselves and others. Most students in a language classroom are unaware
of how they feel about the target language and its speakers, or why it is
important to know and use a second language. Beyond the concrete goals
of learning grammar and vocabulary, they have a largely limited sense of
themselves as language learners and varying perceptions of what language
learning encompasses. Helping students engage in critical reflection is,
however, a somewhat elusive endeavor.
Dörnyei’s (2005, 2009, 2010) work on individual differences among
foreign language learners and the “self-framework” offers a compelling
way to conceptualize an approach to developing students’ awareness. He
proposes a system of the human mind that involves cognition, affect, and
motivation and their dynamic interplay. He states that “Each of the three
mental dimensions can be viewed as dynamic subsystems that have con-
tinuous and complex interaction with each other and which cannot exist
in isolation from one another” (2010, p. 261). Dörnyei uses this tripartite
system to theorize the notion of the “possible self” and notes that “From an

educational perspective the most important possible self is the ‘ideal self’ ”
(2010, p. 265). In Dörnyei’s view, the notion of the “ideal self” is central to
understanding why learners are, or are not, motivated to succeed in
a given endeavor. That is, if a person has a vision of who s/he is and who
s/he wants to become – a possible future self, or an ideal self – the foreign
language learning task becomes integral to that person’s sense of self.
The ideal L2 self is the L2-specific facet of one’s ideal self, which refers
to the
representation of the attributes that someone would ideally like to

possess . . . The theory suggests that if the person we would like to become
speaks an L2, the ideal L2 self is a powerful motivator to learn the L2 because
of the desire to reduce the discrepancy between our actual and ideal selves.
(Dörnyei 2010, p. 257)
Having a sense of self as a multi-competent L2 user does not come about

spontaneously and cannot be assumed in the classroom setting. Research
in psychology confirms that a person’s self-image is highly complex and
that people differ in their abilities to generate a successful sense of self.
Dörnyei (2009) notes that “even if the self image does exist, it may not have
a sufficient degree of elaborateness and vividness to be effective” (p. 19).
Every student in a foreign language classroom has a unique sense of self
and varying ways of envisioning his or her ideal self. The kinds of (utilitar-
ian) tasks that are typical in the foreign language classroom do not inher-
ently promote self-reflection about who students want to be as L2 learners
or users. Therefore, the ideal L2 self must be envisioned in the classroom
and brought into conscious awareness. The multi-competent L2 learner
creates a vision of his/her ideal L2 self; s/he engages in conscious reflection
about multi-competence as opposed to monolingual native speaker-ness.
This multi-competence approach to education places critical reflection
about oneself, one’s own language and culture, and the target language
and culture at the heart of the foreign language curriculum.
21.4 Research on language awareness and multi-

competence
This LAF [Language Awareness Forum] has taught me that learning

French does more than just teach us about the French language or
even about French culture. It teaches the more general ability of
learning how to be multi-competent and how to live in the third
space, an ability that one can then apply to other languages and
cultures. (Student quote from Scott et al. 2013, p. 98)
In an effort to explore the notion of the multi-competent L2 learner who

has an increasing ability to imagine herself as a multi-competent L2 user,
we designed a series of three classroom presentations for novice learners

of French called the Language Awareness Forums (Scott, Dessein, Ledford

and Joseph-Gabriel 2013). The goal of each LAF is to engage students in
discussions about the role language learning plays in their sense of self.
Above all, the aim of these presentations, spaced out over the course of a
fifteen-week semester, is to help students generate a vision of their ideal L2
self in an effort to empower them as lifelong multi-competent L2 users.
The first LAF deals with the topic of language and identity to develop
students’ awareness of the ways learning a second language shapes their
sense of self. The students begin by exploring three guiding questions:
(i) What does “identity” mean? (ii) What are the experiences that have
shaped your identity? (iii) In what ways might your identity change by
studying French? Following both small-group and full-class discussion of
these questions, the teacher engages students in an interactive PowerPoint
presentation about research in the area of identity formation. The focus of
the presentation involves the dynamic and fluid nature of identity; stu-
dents are guided to discuss the idea that identity is not considered to be
fixed or static, but rather as changing during the course of the lifespan.
They explore how studying French makes them confront their precon-
ceived notions about how French-speaking people look, dress, and speak;
they also examine whether they want to speak/act/be like French-speaking
people or whether they want to establish a boundary between themselves
and French-speaking people. Through this discussion, each student is
guided to explore ways his/her sense of self is being shaped by being a
French language learner.
The second LAF explores definitions of bilingualism with the intention
of helping students understand that being bilingual means much more
than having full mastery of two languages. Following the same format as
the first LAF, students work in small groups to discuss three guiding
questions: (i) What does it mean to be bilingual? (ii) Are you bilingual?
(iii) Do you expect to become bilingual by studying French? During the
subsequent teacher-led discussion, students who grew up in families
where languages other than English were spoken are encouraged to talk
about how, when, and where they use their two (or more) languages. It is
especially important for all students to understand that bilingual people
use their two languages differently, depending on the context. This reali-
zation makes them aware that being bilingual does not mean having full
proficiency in two languages. During the PowerPoint presentation,
students learn that there are many different definitions of bilingualism
and that there is no agreement among linguists regarding precisely what
bilingual means. This discussion serves to demystify bilingualism and to
make students aware that no one has the same level of proficiency in two
languages; bilingual people understand, read, write, and speak a second
language to varying degrees. They are introduced to Cook’s notion of the
L2 user who is able to use a language other than his/her native language at
any level for any purpose. In addition, this LAF includes a close analysis – or

deconstruction – of the term “native speaker” to ensure that students

understand that studying French will not make them native speakers of
French. The discussion of what it means to be a native speaker is brought
to a theoretical level by asking students to explore what “native speaker”
means in our twenty-first century world in which people speak varieties of
French (or English, or Spanish, etc.).
The third and final LAF focuses on intercultural awareness and global
citizenship. The three guiding questions are: (i) What would it mean to
develop “intercultural awareness”? (ii) What experiences in your life
have helped you develop intercultural awareness? (iii) What kinds of
learning experiences in our French class might promote intercultural
awareness?
After discussing these questions, students are led to examine
Kramsch’s (1998, 2009a) notion that studying a second language involves
language crossing, or living, speaking, and interacting between spaces,
across multiple languages or varieties of the same language. According to
Kramsch, this language crossing puts learners in a “third space,” or a
privileged position located between one’s own culture and the target
culture. Students are sensitized to the ways in which inhabiting this
privileged third space gives a person a new perspective of oneself and
one’s relation to both cultures. In conjunction with this notion of third
space, this LAF presents the notion of multi-competence which accounts
for a person’s knowledge of his/her first language and a developing
understanding of a second language. Instead of thinking about indivi-
dual languages as separate in the mind of one person, students explore
the ways in which all the languages a person knows interact, mutually
influencing each other. Ultimately students are led to think about the
connections between the notion of third space and the multi-competent
L2 user.
Although we are in the very earliest stages of this research, our preli-
minary findings suggest that the LAFs offer students opportunities to
grapple with ideas that go far beyond the scope of a traditional foreign
language classroom. Some of the most important questions students dis-
cuss during the LAF presentations include:
– In what ways are the languages you speak part of your identity?
– What is the difference between your language and your nationality?
– What is a “native speaker”? Are you a native speaker of a particular
language?
– How does your native language give you a sense of power, or a sense of
belonging to a group?
– Have you ever felt like a “language outsider”?
– What languages are/are not “cool”? Why?
– What languages might give a person access to power?

– Name some languages spoken by people in power. Name some lan-

guages spoken by marginalized people.
– Describe stereotypes associated with particular language groups.
– What languages are stigmatized?
– In what ways can learning a second language give you power?
– What does it mean to be “a global citizen”?
Individual teachers may adapt these questions and include additional

questions that reflect their own areas of interest and concern.
Students’ written reflections about each of the LAFs suggest that they
engage in exploring the relationship between language learning and iden-
tity construction, they deconstruct preconceived notions about what it
means to be bilingual, and they discover that learning French can foster
awareness about language learning that is transferable to an encounter
with any new language or culture (Scott et al. 2013). These initial findings
have been somewhat limited, in part because assessing the desired out-
come, namely a sense of self as multi-competent L2 user, is elusive. We are
continuing to investigate ways to isolate and assess specific dimensions of
the development of an ideal L2 user self. Using an adaptation of the scale
developed by MacIntyre, MacKinnon and Clément (2009), we hope to
further our understanding of how the LAFs change students’ sense of
self, and in particular, how it shapes their understanding of their “ideal
selves” as multi-competent L2 learners. In addition, we are conducting
interviews with students to gain a better understanding of how the LAFs
affect individual students; we have preliminary evidence suggesting that
multilingual students respond quite differently to the LAF presentations
than monolingual students. These interview sessions are informed by
Ushioda’s (2009) notion that individuals must be studied in their particular
contexts. She stresses the importance of
a focus on real persons, rather than on learners as theoretical abstractions;
a focus on the agency of the individual person as a thinking, feeling
human being, with an identity, a personality, a unique history and back-
ground, a person with goals, motives and intentions; a focus on the inter-
action between this self-reflective intentional agent, and the fluid and
complex system of social relations, activities, experiences and multi
micro- and macro-contexts in which the person is embedded, moves,
and is inherently part of. (Ushioda 2009, p. 220)
Ultimately, we believe that the foreign language classroom experience

can affect a person’s ability to imagine a more vivid L2 self. There is still no
clear understanding of how to engage students in this process; however,
we consider the LAFs to be a step in the right direction. We agree with
Dörnyei, who proposes that learners can be motivated by imagining who
they are as language learners: “the first step in a motivational intervention
following the self approach is to help learners construct their Ideal L2 Self,
that is, to create their vision” (2009, p. 33).

21.5 Final thoughts
The purpose is not to replace the familiar with the new,

nor to encourage identification with another culture, but to
de-familiarise and de-centre, so that questions can be raised
about one’s own culturally determined assumptions and
about the society in which one lives. (Byram 2009, p. 199)
Students come to our classrooms with varying levels of competence in the

target language. They may also come to us as multi-competent L2 users of
many different languages. In other words, they may already know and use
two or more languages when they arrive in our classrooms to learn
another one. In general, however, students are not aware of what language
use entails or how the languages in their minds relate, interact, or coop-
erate. Their pre-conceived notions, as well as their teachers’ beliefs about
language learning and teaching, do not typically foster multi-competence.
Multi-competence is a state of mind that exists subconsciously; for it to be
productively operative, multi-competence must be taught, learned,
modeled. The foreign language classroom is the ideal place for this work.
Multi-competent L2 learners are students in foreign language classrooms
where multi-competence is a clearly operating perspective. In this kind
of classroom setting, multi-competent L2 learners are emergent multi-
competent L2 users.
Multi-competence is a state of mind and it affects the whole mind
(Cook, Chapter 1, this volume). Cognition, affect, and motivation are all
fundamental characteristics of the multi-competent mind. Teachers
need to address this whole person, with his/her unique language stories.
The goals of foreign language teaching should include an intentional
focus on how this whole person views him/herself and others in our
multicultural twenty-first century society. If a learner is aware of the
role that language plays in his/her life, and feels that knowing a second
language can be an important part of his/her ideal self, s/he will be
motivated to participate in the community of multi-competent L2
users. Byram (2009) urges us to consider the importance of education
for intercultural citizenship; rather than concentrating on foreign
language study for instrumental purposes, such as for travel or work,
foreign language study should focus on the ways it “can take learners
beyond a focus on their own society, into experience of otherness, or
other cultural and belief values and behaviours” (2009, p. 198). Similarly,
Kramsch states that:
An education based on the transmission of information or on communi-
cative training does not prepare the new generation adequately for the
complexities of a multilingual world. The competencies required to
become a multilingual subject are of a much more symbolic nature than
has been acknowledged up to now. (Kramsch 2009b, p. 190)

Her notion of “symbolic competence” in language education is a reframing

of her notion of “third space” or “third culture” in an effort to stress “a
dynamic, flexible, and more contingent competence” (p. 199). I would
argue that this contingent competence is multi-competence – a
competence that “focuses on the subject position, or the relation
between L1 and L2, C1 and C2, NS and NNS, Us and Them, etc.” (Kramsch
2009b, p. 199).
These are lofty goals, but they are worthy ones. If we agree that the goal
of foreign language learning is not to approximate the native speaker, we
must be able to offer an alternative vision. I believe this vision can be based
on the multi-competence perspective. There is no question that evaluating
this multi-competent L2 learner is elusive; however, drawing from
Dörnyei’s tripartite view of the human mind, we might envision focusing
on three different areas: cognition, affect, and motivation. Although these
three areas are mutually interdependent and cannot readily be isolated
one from the other, assessment might be based on what students know
about the language system (cognition), their emerging feelings about L2
use (affect), and the degree to which they are invested in developing an
ideal L2 self (motivation). If we focus on more than spoken language and
grammatical accuracy, and broaden our understanding of what it means to
be an aware multi-competent L2 user, then we can make room for explor-
ing the ways second language learning shapes one’s sense of self; analyz-
ing when and how people use the target language and the ways it shapes
their values, traditions, and institutions; and, ultimately, recognizing why
this kind of learning is valuable and relevant.
We do not want to turn away from what we are doing in our foreign
language classes. Helping students develop the necessary language skills
to understand, speak, read, and write in the target language is certainly a
worthy goal. It is, however, a limited goal. Those skills will serve students
only if they are taught as part of an overarching goal to promote inter-
cultural awareness. As Byram states:
One of the outcomes of teaching languages (and cultures) should be the
ability to see how different cultures relate to each other – in terms of
similarities and differences – and to act as mediator between them, or
more precisely between people socialised into them. This also includes
“mediating” between oneself and others, i.e. being able to take an “exter-
nal” perspective on oneself as one interacts with others and to analyse
and, where describable, adapt one’s behaviour and underlying values and
beliefs. Thus at any given point in time, the individual is bringing into
contact through their own self, two sets of values, beliefs and behaviours.
(Byram 2008, p. 68)
The ability to learn languages, to become a plurilingual “language person”

with a range of competences to different levels in different languages,
needs to be the focus of language teaching just as much as the ability in a
particular language being taught at a particular time. (Byram 2008, p. 17)

This approach to teaching presumes that being a plurilingual language

person with certain attitudes, knowledge, and skills is a valid goal; it also
presumes that these traits must be learned. Ultimately, this approach is
founded on the notion that learning to think differently will enable a
person to act differently.
Finally, returning to Cook’s house metaphor, a second language lives in
the main house alongside any other languages in the mind of one person.
The foreign language classroom is also part of the main house which
accommodates the greater community of languages and language users.
The foreign language classroom is not the extension at the back, but rather
an integral part of the ecosystem (“eco” from the Greek word oikos meaning
“household”) that survives when its constituents work together to
promote awareness and inclusivity for the betterment of the world.
Given this understanding, a multi-competence perspective is central to
our mission as language educators.
References
ACTFL Proficiency Guidelines 2012. www.actfl.org/sites/default/files/pdfs/

public/ACTFLProficiencyGuidelines2012_FINAL.pdf
Byram, M. 2008. From Foreign Language Education to Education for Intercultural
Citizenship. Clevedon, UK: Multilingual Matters.
Byram, M. 2009. ‘Politics, policies and political action in foreign language
education’, in V. J. Cook and Li Wei (eds), pp. 193–214.
Byram, M. 2012. ‘Language awareness and (critical) cultural awareness: rela-
tionships, comparisons and contrasts’, Language Awareness 21 (1–2), 5–13.
Cook, V. J. 1995. ‘Multi-competence and the learning of many languages’,
Language, Culture and Curriculum 8 (2), 93–98.
Cook, V. J. (ed.) 2002a. Portraits of the L2 User. Clevedon, UK: Multilingual
Matters.
Cook, V. J. 2002b. ‘Background to the L2 user’, in V. J. Cook (ed.), pp. 1–28.
Cook, V. J. 2002c. ‘Language teaching methodology and the L2 user
perspective’, in V. J. Cook (ed.), pp. 327–343.
Cook, V. J. 2005. Multi-competence: black hole or wormhole? Keynote talk
at SLRF, Columbia University. Published as V. J. Cook (2007), ‘Multi-
competence: black hole or wormhole for SLA research?’ in Z.-H. Han
(ed.), Understanding Second Language Process, pp. 16–26, Clevedon:
Cook, V. J. and Li Wei (eds) 2009. Contemporary Applied Linguistics: Language
Teaching and Learning. London: Continuum.
Dörnyei, Z. 2005. The Psychology of the Language Learner: Individual Differences in
Second Language Acquisition. Mahwah, NJ: Lawrence Erlbaum.
Dörnyei, Z. 2009. ‘The L2 motivational self system’ in Z. Dörnyei and
E. Ushioda (eds), pp. 9–42.

Dörnyei, Z. 2010. ‘The relationship between language aptitude and

language learning motivation: individual differences from a dynamic
systems perspective’, in E. Macaro (ed.), pp. 247–267.
Dörnyei, Z. and Ushioda, E. (eds) 2009. Motivation, Language Identity and the L2
Self. Bristol, UK: Multilingual Matters.
Kramsch, C. 1998. ‘The privilege of the intercultural speaker’, in M. Byram
and M. Fleming (eds), Language Learning in Intercultural Perspective:
Approaches through Drama and Ethnography, pp. 16–31. Cambridge
University Press.
Kramsch, C. 2009a. ‘Third culture in language education’, in V. J. Cook and
Li Wei (eds), pp. 233–254.
Kramsch, C. 2009b. The Multilingual Subject. Oxford University Press.
Macaro, E. (ed.) 2010a. The Continuum Companion to Second Language
Acquisition. London: Continuum.
Macaro, E. 2010b. ‘Conclusions’, in E. Macaro (ed.), pp. 300–305.
MacIntyre, P. D., MacKinnon, S. P. and Clément, R. 2009. ‘Toward the
development of a scale to assess possible selves as a source of language
learning motivation’, in Z. Dörnyei and E. Ushioda (eds), pp. 193–214.
Scott, V. M. 2010. Double Talk: Deconstructing Monolingualism in Classroom
Second Language Learning. Upper Saddle River, NJ: Prentice Hall.
Scott, V. M., Dessein, E., Ledford, J. and Joseph-Gabriel, A. 2013. ‘Language
awareness in the French classroom’, The French Review 86, 90–102.
Svalberg, A. 2007. ‘Language awareness and language learning’, Language
Teaching 40 (4), 287–308.
Ushioda, E. 2009. ‘A person-in-context relational view of emergent
motivation, self and identity’, in Z. Dörnyei and E. Ushioda (eds),
pp. 215–228.

22
Multi-competence
and emotion
Jean-Marc Dewaele
22.1 Introduction
Cook (2012) defines multi-competence as ‘the knowledge of more than

one language in the same mind or the same community’ (p. 3768), by
which he means ‘everything a single person or a single community
knows about all the languages they use’ (p. 3768). Reading the definition
one might be forgiven for thinking that this is not about emotion. Indeed:
‘Multi-competence therefore involves the whole mind of the speaker, not
simply their first language (L1) or their second’ (p. 3768). However, Cook
(personal communication) clarified that emotions are presumably
located in the mind and that his use of ‘mind’ was supposed to include
emotions and anything else ‘mental’. An important aspect of the multi-
competence perspective is the positive outlook: rather than focusing on
deficits in the languages, it investigates the deviations and celebrates the
intricate interactions between the languages.
Pavlenko (2003, p. 58) concluded her chapter with a very apt quote by the
Polish–English writer Eva Hoffman:
When I speak Polish now, it is infiltrated, permeated, and inflected by the
English in my head. Each language modifies the other, crossbreeds with it,
fertilises it. Each language makes the other relative. (Hoffman 1989, p. 273)
While Cook’s perspective is resolutely cognitive, he has not excluded

enquiries outside this domain, thus agreeing that the acquisition of a
second language (L2) can have consequences that are not strictly linguistic:
‘Acquiring another language alters the L2 user’s mind in ways that go
beyond the actual knowledge of language itself’ (Cook 2002, p. 7).
One aspect that has gradually been taken up by applied linguists
(Herdina and Jessner 2002; Schmid 2011) is the idea that multi-competence
I would like to thank Aneta Pavlenko for her feedback on parts of this chapter.

is highly dynamic and in a constant state of flux. In Dewaele and Pavlenko

(2003) we argued that an individual’s multi-competence is not a fixed
end-state.1 There is constant variation both within and between multi-
linguals. It is highly unlikely that two multilinguals would ever have
isomorphic multi-competence (the same is true for monolinguals) as
subtle differences in linguistic input and output could have consequences
for individuals’ multi-competence. We used a paint metaphor to compare
the two languages in contact in the multilingual’s mind to:
two liquid colours that blend unevenly, i.e. some areas will take on the
new colour resulting from the mixing, but other areas will retain the
original colour while still others may look like the new colour, but a closer
look may reveal a slightly different hue according to the viewer’s angle.
(Dewaele and Pavlenko 2003, p. 137)
We thus defended a view of multi-competence as an ever-changing and

highly complex system. Parts of the system can be in equilibrium for a
while, but an unexpected change in the internal and external environ-
ment, i.e. a change in the frequency or nature of the linguistic input, or a
specific linguistic activity – such as reading a book or watching a film with
an unfamiliar sociolect – can cause widespread restructuring, with some
‘islands’ remaining in their original state (Dewaele and Pavlenko 2003).
This view is particularly important in considering emotion, because of
the inherent variability in emotional responses between people in similar
contexts. People from the same speech community might find some emo-
tional expressions uttered by one member of the community harmless
but could perceive emotionally loaded expressions or terms uttered by
another member of the community as offensive. A typical example is
the use of racial terms that are acceptable from fellow members of the
community, but deemed racist when used by outsiders. Even a single
individual might react differently at different moments in time depending
on mood, tiredness or state of inebriation, and be judged differently by the
people around him/her.
Moreover, emotions are displayed very differently depending on the
occasion: ‘one may turn red with anger, glower and shout in one situation
and appear white-faced, expressionless and icily polite in another’
(Wierzbicka and Harkins 2001, p. 2). A lot depends on the degree of self-
control of the individual; it may be harder to control involuntary signs of
emotion compared to the conscious uttering of emotional speech.
We have argued that the learning of multiple languages, and the result-
ing multi-competence, affects not just phonology, morphology, syntax,
lexical choices, but also pragmatics, where the communication of emotion
is situated (Dewaele 2013; Pavlenko 2005). Just as Cook and Bassetti (2011)
argue that L2 users think differently from monolinguals, we would also
argue that they express their feelings differently. In other words, the fact
of learning to recognise and express emotions in different languages

Multi-competence and emotion 463
affects how these emotions are recognised and expressed in the multi-
lingual’s various languages. Again, L2 users have been shown to have a
slightly different linguistic knowledge of their L1 (Cook and Bassetti 2011);
they have probably also shifted a little in their ability to recognise and
express emotions.
Section 22.2 will briefly discuss the difficult distinction between
effects of language and culture on conceptual change in bi- and multi-
linguals. After that, in Section 22.3, we will present an overview of
research that combined emotion and multilingualism research, some-
times with explicit references to the multi-competence perspective.
We will look more specifically at multi-competence in the grammar
and lexicon of emotion, and in language choices to express emotion.
Then, in Section 22.4, we will focus on emotional acculturation. Finally,
Section 22.5 will consider the emotional range of multi-competent
individuals and the resulting hybrid sense of self.
22.2 What is the effect of language, what is the effect

of culture?
While the multi-competence perspective focuses on the effect of the new

language rather than the new culture on the mind of the speaker, it is in
fact quite difficult to separate these two entwined variables. They can be
considered separate sides of the same coin. That does not stop some
politicians from trying to filter out the cultural element in foreign lan-
guage teaching in order to preserve national cultural loyalty: English is
considered to be language of the enemy in many countries, with ‘ridicu-
lous’ ideas about personal freedom and democracy. Yet, the need for
English is such that authorities have to organise its teaching, but in such
a way that no cultural–ideological ‘contamination’ can occur. This does
pose a problem, because knowing a language involves more than purely
linguistic knowledge. However, this probably worries the authorities less
than having their students create imaginary alternative identities for
themselves (Lvovich 1997).
While new cultural knowledge might lag behind the acquisition of new
linguistic knowledge in the first stages of foreign language (FL) learning in
a context of formal instruction, cultural knowledge will gradually seep in
through contact with newspapers, books and films in the FL, and with
speakers of that language. Pavlenko (1999) offered a nice example of this in
her study on conceptual differences between Russian and American
English in the notions of privacy and personal space. Monolingual
Americans mentioned these notions in retelling silent films that showed
a violation of privacy and personal space, something Russian monolin-
guals did not mention. She repeated the experiment with two groups of
Russian–English bilinguals, a group living in the US who had plenty of

cultural knowledge, and a group of Russian FL learners of English who

had much less cultural knowledge or awareness. While the former group
mentioned the invasion of privacy and violation of personal space, both in
English and – strikingly – in Russian, the latter group did not mention it,
despite knowing the meaning of the English word privacy. Pavlenko (1999)
pointed out that semantic representations (defined as ‘explicitly available
information on the meaning of various forms’) need to be distinguished
from conceptual representations (defined as ‘multi-modal mental repre-
sentations’: p. 212), and that the latter typically require more time to
develop. One example of this is my own conceptual representation of
‘horse’: in French it is cheval, in Dutch paard – which seem like perfect
translation equivalents of the English word. I became aware of the very
different conceptual representation British people have of ‘horse’ when
meat lasagnes sold in British shops were found to contain horsemeat. It
provoked a national scandal, as if some form of cannibalism was involved.
The French and Dutch concept of ‘horse’ brings to mind the gentle animal
that allows rich people to ride on, but also one of the possible meats you
can order at your local butcher. In other words, I realised that eating
horsemeat is a taboo for the British because of cultural values embedded
in their conceptual representation of that animal. It would therefore be
interesting to see whether the conceptual representation of ‘horse’ among
British people in horse-eating countries has evolved.
22.3 Multi-competence and the expression of emotion
22.3.1 Grammar
Pavlenko (2008) and Pavlenko and Driagina (2007) have investigated the
structural non-equivalence of the expression of emotion in English and
Russian, and the consequences that this has on users of these languages.
While both languages have emotion nouns, adjectives, transitive and
intransitive verbs and have comparable morphosyntactic categories,
the distribution is quite different (Pavlenko and Driagina 2007, p. 215).
Russian has more intransitive emotion verbs. The pattern of emotion
coding is also quite different in both languages, with L1 users of English
favouring adjectival constructions in combination with copula verbs to
express emotions, while L1 users of Russian favour verbs, specifically
intransitive and reflexive emotion verbs (p. 215).
The analysis of retellings of short silent films mentioned earlier showed
that the L1 English speakers favoured emotion adjectives such as upset
(75% of emotion word tokens) whereas the Russian L1 speakers preferred
emotion verbs (51% of emotion word tokens), such as rasstroit’sia ‘to get
upset’ (Pavlenko 2002, p. 67). Advanced North American learners trans-
ferred the L1 preference to their Russian L2, resulting in negative transfer,
for example become with emotion adjectives such as stala serditoi ‘became

angry’, while native speakers of Russian consistently used emotion verbs,

saying ona rasserdilas ‘she got angry’ (Pavlenko 2008, p. 153). Interestingly,
from a multi-competence point of view, Pavlenko (2003) found that 21
Russian native speakers who had English as an L2, and who had been living
in the US for less than 10 years, exhibited extensive L2 influence on the L1.
As some bilinguals retold the films in Russian, they framed emotions
linguistically as states, rather than as active processes, violating both
semantic and syntactic constraints of Russian.
22.3.2 Lexicon
Pavlenko and Driagina (2007) also considered how the American learners
of Russian, and the native speakers of English and Russian, use emotion
words that lack translation equivalents in the other language – for exam-
ple, the Russian verb perezhivat, which refers to ‘the process of worrying,
taking things hard and experiencing them keenly, or, literally, suffering
things through’ (p. 217). In turn, common English emotion words as fun or
frustration have no single-word equivalents in Russian (p. 217). The authors
found that the American learners did not use perezhivat a single time in
their film retellings, despite admitting in post-experiment debriefings to
having encountered this verb before (p. 226).
The Russian bilinguals (with L2 English) in Pavlenko (2003) were aware
that their exposure to English had affected their ability to express their
emotions in Russian. One participant complained about the lack of a
Russian translation of the English concept privacy (p. 54), another used
the English word clumsy in describing how she felt using Russian: ‘I feel
more and more clumsy, uncomfortable, when I speak Russian’ (p. 55).
Strikingly, only one bilingual participant used the Russian word perezhivat,
which dominated the narratives of Russian monolinguals (p. 56). It thus
seems that intense exposure to the L2, and to L2 culture, leads to a
‘conceptual restructuring’ of emotion concepts in the L1, namely
‘readjustment of the category structure and boundaries in accordance
with the constraints of the target linguistic category’ as well as ‘conceptual
development’ defined as ‘development of new multimodal representa-
tions that allows speakers to map new words onto real-world referents
similar to native speakers of the target language’ (Pavlenko 2009, p. 141).
Pavlenko (2003) concluded that ‘the complex phenomenon of L2 influence
on L1 is best understood from a multi-competence perspective’ (p. 58).
Another major contribution in this area of research is the work by
Stepanova Sachs and Coley (2006), who compared representations of
Russian and English translation equivalents in 22 English and 22 Russian
monolinguals and 22 Russian–English bilinguals of the pair jealousy/revnost
and envy/zavist. In English the words jealous and envious can be used as
overlapping, as synonyms, on many occasions, but in Russian their trans-
lation equivalents cannot because they are categorically distinct. The

Russian nouns revnost and zavist go with distinct adjectives, and Russian
monolinguals favour the emotion verbs revnuet and zaviduet. Participants
had to select a word to describe a jealousy or an envy story they had heard.
The English monolinguals considered the words envious and jealous as being
equally appropriate for describing the emotions of characters in envy
stories. Bilinguals tested in Russian distinguished jealousy and envy
stories, whereas they made no distinction for envy stories in English,
just like native English speakers (p. 226). The authors found a greater
differentiation between revnuet and zaviduet in envy stories for Russian
monolinguals than for bilinguals tested in Russian. It thus seems that the
advanced knowledge of L2 English had blurred the bilinguals’ category
boundary between revnuet and zaviduet in their L1. In a second experiment,
involving a free sorting task, English monolinguals and bilinguals were
more likely to group envy and jealousy situations together than were
Russian monolinguals.
Finally, Ożańska-Ponikwia (2013) investigated changes in the emotional
repertoire of 102 Polish L2 users who had lived, or were living, in English-
speaking countries. She found that L2 socialisation had influenced the
bilinguals’ perception of the specific Polish emotion of tesknota (‘yearn-
ing/longing/Sehnsucht’). Participants had to describe the emotion felt by
the main character in two short stories in Polish and English. A control
group of Poles, with English as L2, who had never lived outside Poland all
mentioned tesknota. Three-quarters of the Polish–English bilinguals of the
experimental group produced tesknota in the Polish version of the story;
the others used other Polish emotion words or English emotion words
(nostalgia, anxiety). After reading the English version of the story, a third
of the bilinguals wrote the English words loneliness, followed by sadness,
homesickness, longing, with only 7% producing tesknota. Ożańska-Ponikwia
thus concluded that their L2 socialisation had blurred their perception of
this specific Polish emotion. To sum up, there is clear evidence of concep-
tual restructuring in the expression of emotion among multi-competent
speakers, both in their grammar and their lexicon.
22.3.3 Language choices

The first large-scale investigation into language choices of multilingual
and multicultural individuals was based on data collected through the
Bilingualism and Emotions Questionnaire (BEQ) (Dewaele and Pavlenko
2001–2003) from more than 1500 multilinguals (Dewaele 2013).
Multilinguals’ dominant language, typically their L1, was the preferred
language to express emotions. However, other languages were sometimes
used to express emotion. This could be interpreted as evidence of multi-
competence, as these multilinguals were no longer restricted to a single
channel to express their emotions. Participants who had learned a foreign
language (LX, which could be an L2, L3, L4 or L5) through classroom

instruction and had simultaneously used that language to communicate

outside the classroom, and participants who were early starters in the
acquisition of the LX tended to use that language more frequently to
express emotion than participants who had purely formal instruction
and were later starters. An LX that was used frequently was typically also
used frequently to express emotion. An analysis of individual variation in
the perceived emotional force of swearwords in the multilinguals’ differ-
ent languages and the perceived emotional weight of the phrase ‘I love
you’ revealed similar patterns (Dewaele 2004, 2008, 2013).
Balanced bi- or multilinguals – reporting shared dominance in two
languages – were not using their languages in equal measure to express
emotions (Dewaele 2011). The 386 multilinguals from the BEQ with max-
imal proficiency in an L1 and an LX, who used both languages constantly,
did prefer the L1 for communicating feelings and anger, and for swearing.
Participants reported that their L1 was emotionally stronger than the LX
and that they felt less anxious in their L1. The analysis of interviews
between the interviewer (Benedetta Bassetti) and 20 participants on the
topics covered by the BEQ revealed that a longer stay in the LX culture was
linked to a gradual shift in language preferences and perceptions, with the
LX starting to match the L1 in hearts and minds. It also emerged that LX
linguistic practices seeped into L1 exchanges with interlocutors sharing
the same linguistic profile. Interviewees’ views on swearing in their dif-
ferent languages drew particularly rich responses, possibly because it is
felt to be a relatively ‘safer’ topic of conversation than display of anger or
of romantic feelings.
Michelle (Taiwanese L1, Mandarin L2, English L3), a Chinese Londoner,
married to an Englishman, reported equal fluency in English and
Mandarin. She reported that, despite the fact that Chinese sociocultural
norms forbade her from swearing, she did use mild English swearwords
in interactions in Mandarin with her Chinese friends in London:
B: Do you ever use Chinese with your husband?

M I : No! not even I swear or, because I don’t swear in
Chinese you see.
B: I see yeah.
M I : It’s not because I’m good, it’s just because education, you
see ehm it’s different, ehm, English swearing is different
from Chinese swearing. English swearing is quite common
to even, you know, whatever you educated whatever you
are, you do it, but in Chinese you really, most educated
people don’t swear.
B: So do feel something is missing when you speak Chinese,
because you can’t swear?
M I : Ah ah! I haven’t thought about that!
B: [laughs]

M I : Maybe, yes, maybe, no. It’s funny, you do get by isn’t it

without swearing, you still get by, but I just think that even
now I swear, I swear when I’m with my friends, Chinese
friends, you have to say ‘oh shoot’ or ‘sugar’ or whatever,
and you know and then you say that in English, so . . .
B: While you speak in Chinese?
M I : Yeah.
B: While you speak in Chinese you swear in English?
M I : Yeah.
B: Or while you speak in English?
M I : While I speak in Chinese, both. I never, I still swear
because again I think it’s a habit, because that reaction
just come out, so it’s a bit like you have to ask yourself
what do I do before I knew English swearing, ehm, how
do I survive?
(Dewaele 2013, pp. 119–120)
Layla (Tunisian Arabic L1, English L2), who feels equally proficient in
both languages and has lived for five years in English-speaking countries,
reports that swearing is taboo in her L1 culture, but that she might use
mild swearwords in English:
L: Speaking of swearing, . . . I never swear in Arabic, never never

at all, because I know exactly what it means, because it’s my
language anyway, and how offensive it would be to swear, but
in English because it’s not my native language, sometimes
I use some swearwords, but I don’t really aware I’m not really
aware of how immense those words. One of the words that
sometimes I use is ‘bloody’, ‘bloody rude’ you know, this is
the only swearword I use.
B: The only one? I see
L: Yeah. Sometimes I use another one, you want to know what,
you want to hear what kind of words I use?
B: Yes
L: Sometimes I say ‘shit’.
B: Ok, right
L: But it’s just because I know many, many of my friends they
keep using it, and I’m not really aware of how immense and
strong this word, because it’s not my native language so I feel
like, I feel much more confident in swearing in English than
Arabic
B: But if you’re speaking Arabic and you are in a situation where
you would normally use the s-word in English, what do you
say in Arabic?
L: I don’t, I don’t because it’s just, I was raised up in, my family
never swear, . . . and I never got used to it, and basically it’s

part of my, not only how I raised, it’s also part of my beliefs
that I don’t like to swear because I think it’s uncivilised, it’s
uncivilised way of speaking, and I feel that I can use any,
although sometimes you really really feel you’d like to do it,
but I don’t in Arabic, I never never say any swearword in
Arabic, I never really honestly.
(Dewaele 2013, pp. 124–125)
The effect of strong socialisation in L2 English has an effect on

language preferences to express anger among Japanese returnees.
It allows multilinguals to vent their anger in a socially accepta-
ble way:
Ryoko (Japanese L1, English L2): I tend to use English when I am angry,
Japanese when I’m hurt or sad, both when I am happy or excited . . . My
other bilingual friends who are all returnees like me said the same
thing about using English when they’re angry. I guess I like the sound
of the swearing words since I heard it so many times during my stay in
the US. This swearing doesn’t happen so often in Japan. It’s a cultural
difference. (Dewaele 2013, p. 122)
Another Japanese participant, Miho (Japanese L1, English L2, Thai L3,
German L4, dominant in L1 and L2) expresses strong emotions in
English with bilingual interlocutors, but sticks to English or Japanese
with monolingual interlocutors. She looked surprised when asked how
she expresses her anger in Japanese and explains that she does it
without words:
B: You’re angry at a Japanese friend who doesn’t understand

English, which language do you use?
M: Um, Japanese.
B: Ah-ah.
M: But I don’t know how to say.
B: So what do you say?
M: I just show angry face?
B: Ah ah.
M: Yeah.
(Dewaele 2013, p. 122)
Tomomi (Japanese L1, English L2, Italian L3, Spanish L4, married to an
Italian, dominant in Japanese and living in the UK for four years) mentions
L1 sociocultural constraints weighing on the use of Japanese swearwords
in Japan. When asked about language preferences to express anger, she
explains that she prefers Japanese when alone, Japanese or English in
written communication with bilingual addressees, because she can
express her feelings more clearly (p. 209). Reflecting on the advantages of
Japanese and English, she mentions that Japanese goes deeper but that
English is straighter:

T : if I wanna express more deep then it’s better in Japanese, but

I feel like Japanese language is not really straight, so if I
wanna show really anger to somebody it’s much better in
English actually.
(Dewaele 2013, pp. 209–210)
She claims that she does not swear in any language:
B: Do you swear?
T: I don’t.
B: No, never, in any language?
T: No, in Japanese we don’t really have swearword, and English
swearword I don’t like it, especially you know with the kids,
they can get so easily, so I don’t have this habit to use swear-
word so I don’t.
B : Have you learnt any Italian swearwords?
T : I know it because you know some people around me say that,
but I don’t, I don’t use any.
(Dewaele 2013, p. 124)
Quipinia (Cantonese L1, English L2), living in Hong Kong, reported that
her parents frowned upon the expression of emotion, ‘therefore I feel a lot
easier to use another language to express the feelings and the different
personality inside me’ (Dewaele 2013, pp. 122–123). She remembers an
argument in which she burst out in English at her parents who are profi-
cient in English but with whom she usually uses Cantonese:
Quipinia: But I remember one time when they were arguing with me and
I was soooooooooo angry that I shouted out ‘IT’S UNFAIR!!!!’ I guess it’s
regarded quite impolite if I shouted at my parents (you know Chinese
Traditional family) but at that point I feel that I had to express my anger
and let myself just do it in another language; perhaps I feel I’m another
person if I say that in English. (Dewaele 2013, p. 123)
Veronica Zhengdao Ye, a Chinese scholar who emigrated to Australia at

the age of 23, and whose 2004 paper will be discussed in the next section,
also reported that using swearwords in Chinese L1 is impossible for her,
but that using English swearwords in the Australian context is perfectly
acceptable, although she realises she might be overdoing it:
I belong to the group of people who are brought up with the notion that
swearing is uncivil. And I have NEVER used swear and taboo words in my
L1. But I do use words in English which native English speakers would
consider uncivil to use, such as ‘shit’ and ‘pissed off’. I could use them
exactly because I do not have the same sense of emotional weight of these
words as do the native people. My only clue of how ‘strong’ these words
are was from people’s reaction when I used them. My friends are often
astonished when I use them, because they say that I do not look like the
person who could say those words. When I use them, they say that they

knew I feel something very strong about something. I myself don’t

mind using those English swear words, as when I use them, I have fun of
being another person for a moment!
(Veronica Zhengdao Ye, personal communication, September 2004)
Several participants, typically of Asian or Arabic origin, thus reported
using an LX to express emotions, including some swearwords, to escape
the L1 social and cultural constraints. Contact with English language and
culture had made them more inclined to express anger in the English way,
even when communicating in the L1. This could include the use of mild
swearwords. There was a considerable amount of variation both between
and within individuals: some stuck to non-verbal communication of anger
in interactions with monolingual Japanese, many expressed a different
language preference depending on the identity of the interlocutor/s. Some
liked the directness of expressing anger in English, but rejected swearing
in that language. All had spent a considerable amount of time in an
English-speaking environment and were highly proficient in English.
They could probably be described as multi-competent biculturals who
had reflected on the differences between their L1 norms and the English
norms and had thus developed a good amount of meta-pragmatic aware-
ness. Those who did not use English systematically within the family
reported language preferences and swearing practices situated halfway
between L1 norms and English LX norms. Layla and Michelle reported
using weak English swearwords, and in doing so, clearly distinguished
themselves from their L1 monolingual and monocultural peers. Veronica
Zhengdao Ye on the other hand, seemed to be enjoying the transgression
of her L1 rules, in swearing like her fellow Australians.
22.4 Emotional acculturation
Veronica Zhengdao Ye refers in the title of her 2004 paper La Double Vie de
Veronica to the brilliant and dramatic 1991 French- and Polish-language
film by Krzysztof Kieslowski, La Double Vie de Véronique. The main charac-
ters are Weronika, a Polish choir soprano, and her double, Véronique, a
French music teacher, both interpreted by Irène Jacob. Although the
two women do not know each other, they share a mysterious emotional
bond that transcends language and country. Ye explains in a footnote that
the film title and topic was appropriate for the account of her double life
as a Chinese and English bilingual living within two cultures. It reflected
her own travels to and fro between English, Mandarin Chinese and
Shanghainese, her mother tongue. The relationship between the lan-
guages is a constant struggle: ‘When speaking English, I may think in
English, but only partially; the next moment, it flicks back to Chinese.
Sometimes I get confused and the two languages merge – one on top of the
other’ (Ye 2004, p. 138). She prefers the Chinese way of expressing

emotions: ‘subtle, implicit and without words’ (pp. 139–140). She describes
her first parting from her parents, just before boarding the plane to
Australia:
we fought back our tears and urged each other repeatedly to take care; we
wore the biggest smiles to wave good-bye to each other, to soothe each
others’ worries. Just like any other Chinese parting between those who
love each other – there were no hugs and no ‘I love you’. Yet I have never
doubted my parents’ profound love for me. (Ye 2004, pp. 140–142)
She explains that at the beginning of her stay in Australia, she felt uncom-
fortable talking about her true feelings, it made her inner self feel ‘stripped
and vulnerable’. She was struck by the ease with which Australians use
emotion words which made her blush. She gradually understood that
these expressions are niceties for social purposes. She needed some time
before she was able to recognise the emotions displayed in the Australian
context accurately and deal with them appropriately. Interestingly,
two years later, on the plane home to attend her father’s funeral,
she deeply regrets never having hugged him, and decides to give her
mother a big long hug ‘to abridge the physical separation’. Ye’s experience
could be described as the development of linguistic and emotional multi-
competence, and as the typical emotional acculturation of an immigrant.
De Leersnyder, Mesquita and Kim (2011, pp. 452–462) define it as ‘changes
in emotional patterns due to an immigrant’s exposure to and contact with
a new or second cultural context’ (p. 452). The authors point out that the
emotional experiences of people who live together (dyads, groups, cul-
tures) tend to be similar and that immigrants’ emotions probably approxi-
mate host culture patterns of emotional experience. Although they do not
refer to multi-competence – as they are not specifically interested in
language – there are some striking similarities. Indeed, the emotional
experience of the immigrants shifts as a result of the contact with inhabi-
tants of the host culture. The authors carried out a first study on 47 Korean
immigrants and 44 European Americans in the United States and a second
study on 59 first- and 85 second-generation Turkish immigrants as well as
83 Flemish Belgians living in Belgium. They used the Emotional Patterns
Questionnaire to collect data on emotional experiences of immigrants and
host group members. They calculated differences in emotional patterns
using comparable emotional situations. Participants were asked to
describe an event from their own daily life that met the description of an
emotional situation in a particular prompt. No significant differences
emerged in reported emotional events across cultural groups nor across
acculturation levels. However, ‘patterns of emotional experiences differed
in ways that may be considered evidence for emotional acculturation’
(p. 460). The degree of immigrants’ emotional similarity to the host
group was reflected in a correlation value of their individual emotional
patterns with that of the average pattern of the host group. Immigrants’

exposure to and engagement in the host culture predicted emotional

acculturation. In other words, immigrants who had spent a larger propor-
tion of their life in the host country were more likely to have emotionally
acculturated as a result of intercultural interactions and relationships.
Interestingly, emotional concordance was found to be higher for the
positive than for the negative situations. It may thus be easier for an
immigrant to learn, and adapt to, the host culture’s emotional pattern in
positive rather than in negative emotional situations.
The authors raise a question about the changes that underlie the shifts in
emotional patterns: is it because immigrants in the new culture experi-
ence different situations or because they start appraising the same situa-
tions differently? Finally, emotional concordance with the host culture
was not linked to immigrants’ attitudes toward the adoption of values and
traditions, nor to social relationships with members of the mainstream. In
other words, emotional acculturation seems unrelated to the attitudes
toward acculturation (i.e. explicit beliefs) (de Leersnyder, Mesquita and
Kim 2011, p. 462).
22.5 Are multi-competent people multi-emotional?
One of the themes that emerged from previous sections was the extension
of the conceptual and emotional range that multilinguals experienced. In
other words, their multi-competence was linked to a new-found capacity,
and freedom to express emotions, typically through code-switching, that
they would not have expressed as monolinguals. This might lead to a
different perception of self in the various languages, as Ye exclaimed:
‘I have fun of being another person for a moment!’
Pavlenko (2006b) analysed the feedback from 1039 participants of the
BEQ (Dewaele and Pavlenko 2001–2003) and found that almost two-thirds
of participants reported that they feel like different people when they
change languages, a quarter of participants felt no difference, with the
remaining 10% of participants giving no clear response (Pavlenko 2006b,
p. 10). Many participants answered that they felt more ‘real’ and ‘natural’
in their L1, and more ‘fake’, ‘artificial’ in later learned languages (p. 18).
This finding was confirmed in Dewaele and Nakano (2012), where 106
multilinguals reported feeling significantly less logical, less serious, less
emotional and increasingly fake when using the L2, L3 and L4 compared to
their L1.
Wilson (2013) investigated the positive feelings about using foreign
languages among 172 adult first-language English speakers learning an
LX for pleasure or using it in a social setting with other L1 English speakers.
Several LX users reported that operating in the LX gave them a sense of
freedom and enabled them to speak and behave in ways that were differ-
ent from their usual modes.

Ożańska-Ponikwia (2012) studied self-reported feelings of differences

linked to using an L2 among 102 Polish L2 users of English who had
never been abroad and Polish–English bilinguals who had spent some
time in an English-speaking country. She found that self-perceived
changes in behaviour when using the L2 were linked to the personality
traits extraversion, agreeableness, openness, emotion expression, empathy, social
awareness, emotion perception, emotion management, emotionality and sociability.
She argues that participants who scored high on these dimensions were
more likely to be aware of changes in their behaviour when switching
languages.
A pioneer in this area of research, Koven (1998, 2001, 2006), building on
the work of Ervin-Tripp (1967) with Japanese–English bilingual women
married to American men, asked two French–Portuguese bilinguals to tell
the same story in both languages and subsequently interviewed them
about how they felt while telling the story. She found that both partici-
pants: ‘perform(ed), enact(ed), or inhabit(ed) the role of their characters in
the stories quite differently . . . Isabel sounds like an angry, hip suburbanite
in French, whereas in Portuguese, she seems a frustrated, but patient, well-
mannered bank customer who does not want to draw attention to the fact
that she is an émigré’ (Koven 1998, p. 435). The two women seemed to let
themselves be pushed around more when they spoke Portuguese and
stood up for themselves more when they spoke French (Koven 1998). The
different languages allowed the women to ‘perform a variety of cultural
selves’ (Koven 2001, p. 513). Koven focused specifically on Linda, who was
asked to tell stories about the same bad experience in Portuguese and in
French, to a Portuguese–French bilingual of her own age (Koven 2006).
A formal analysis of interlocutory devices and different styles suggested
that she was ‘angrier, more forceful and more aggressive in French’
(p. 107). Koven notes that Linda is aware that her lack of profane or vulgar
vocabulary in Portuguese affects her presentation in that language: ‘Linda
may not be free to perform an aggressive persona in Portuguese’ (p. 108).
22.6 Conclusion
The overview of research of multilingualism and emotion from a multi-

competence perspective leads to a number of general observations. The
addition of a language (or a culture) to an individual’s repertoire has profound
repercussions on the whole system, including the individual’s emotional
geography. The acquisition of new emotion concepts, and of new socioprag-
matic and sociocultural information governing their use, affects the L1
emotion concepts and the way they are verbalised. These effects include
unconscious blurring of categories, but can also trigger a highly conscious
progression into new emotional territory. The L1 concepts, values and
practices remain important for these multilinguals. They typically rate

their L1 as being more emotional than their LX. However, there are clear
instances of blending of L1 and LX values and practices. While they may
be able to keep their languages apart in interactions, more permeability
develops between the two languages at a conceptual level. LX affective
socialisation results in a unique multi-competent behaviour both in the L1
and LX. Swearing in the LX illustrates the newfound freedom to express
oneself without violating L1 norms. The new language and culture offer LX
users new potential emotional selves which they can deploy according
to their needs. Hoffman (1989) referred to cross-fertilisation between her
languages. Metaphorically multi-competence could be illustrated through a
garden metaphor where emotion concepts in language 1 are laid out in neat
separate patches of red flowers (think of a formal French garden), and where
the acquisition of emotion concepts in language 2 or 3 lead to the emergence
of blue and orange flowers between the red flowers in slightly different
patches. As a consequence, red flowers start appearing where they did not
grow before, intermingling with the patches of blue and orange flowers
which also develop their own unique shapes and hues (in other words, an
English herbaceous border).
Note
1. As one reviewer pointed out, the same is true of competence – an

individual’s knowledge of language at any stage of development, i.e. a
five-year-old’s competence which is evolving daily.
References

De Leersnyder, J., Mesquita, B. and Kim, H. 2011. ‘Where do my emotions
belong? A study on immigrants’ emotional acculturation’, Personality and
Social Psychology Bulletin 37, 451–463.
Dewaele, J.-M. 2004. ‘The emotional force of swearwords and taboo words
in the speech of multilinguals’, Journal of Multilingual and Multicultural
Dewaele, J.-M. 2008. ‘The emotional weight of “I love you” in multi-
lingual’s languages’, Journal of Pragmatics 40, 1753–1780.
Dewaele, J.-M. 2011. ‘The differences in self-reported use and perception of
the L1 and L2 of maximally proficient bi- and multilinguals: a

quantitative and qualitative investigation’, International Journal of the

Dewaele, J.-M. 2013. Emotions in Multiple Languages, second edition.
Basingstoke: Palgrave Macmillan.
Dewaele, J.-M. and Nakano, S. 2012. ‘Multilinguals’ perceptions of feeling
different when switching languages’, Journal of Multilingual and
Dewaele, J.-M. and Pavlenko, A. 2003. ‘Productivity and lexical diversity
in native and non-native speech: a study of cross-cultural effects’, in
V. J. Cook (ed.), The Effects of the Second Language on the First, pp. 120–141.
Ervin-Tripp, S. M. 1967. ‘An Issei learns English’, Journal of Social Issues 23 (2),
78–90.
Perspectives of Change in Psycholinguistics. Clevedon: Multilingual
Matters.
Hoffman, E. 1989. Lost in Translation: A Life in a New Language. Vintage:
London.
Koven, M. 1998. ‘Two languages in the Self/The Self in two languages:
French–Portuguese verbal enactments and experiences of self in narra-
tive discourse’, Ethos 26, 410–455.
Koven, M. 2001. ‘Comparing bilinguals’ quoted performance of self and
others in telling the same experiences in two languages’, Language in
Society 30, 513–558.
Koven, M. 2006. ‘Feeling in two languages: a comparative analysis of a
bilingual’s affective displays in French and Portuguese’, in A. Pavlenko
(ed.), pp. 84–117.
Lvovich, N. 1997. The Multilingual Self: An Inquiry into Language Learning.
Ożańska-Ponikwia, K. 2012. ‘What has personality and emotional
intelligence to do with “feeling different” while using a foreign
language?’, International Journal of Bilingual Education and Bilingualism
15, 217–234.
Ożańska-Ponikwia, K. 2013. Emotions from a Bilingual Point of View: Personality
and Emotional Intelligence in Relation to Perception and Expression of Emotions in
the L1 and L2. Newcastle upon Tyne: Cambridge Scholars.
Pavlenko, A. 1999. ‘New approaches to concepts in bilingual memory’,
Pavlenko, A. 2002. ‘Bilingualism and emotions’, Multilingua 21, 45–78.
Pavlenko, A. (ed.) 2006a. Bilingual Minds: Emotional Experience, Expression, and
Representation. Clevedon: Multilingual Matters
Pavlenko, A. 2006b. ‘Bilingual selves’, in A. Pavlenko (ed.), pp. 1–33.

Pavlenko, A. 2008. ‘Emotion and emotion-laden words in the bilingual

lexicon’, Bilingualism: Language and Cognition 11, 147–164.
Mental Lexicon: Interdisciplinary Approaches, pp. 125–160. Clevedon, UK:
Pavlenko, A. and Driagina, V. 2007. ‘Russian emotion vocabulary in
American learners’ narratives’, Modern Language Journal 91, 213–234.
Schmid, M. S. 2011. ‘Contact x time: external factors and variability in L1
attrition’, in M. S. Schmid and W. Lowie (eds), Modelling Bilingualism: From
Structure to Chaos. In Honor of Kees de Bot, pp. 155–176. Amsterdam:
Benjamins.
Stepanova Sachs, O. and Coley, J. D. 2006. ‘Envy and jealousy in Russian
and English: labeling and conceptualization of emotions by monolin-
guals and bilinguals’, in A. Pavlenko (ed.), pp. 209–231.
Wierzbicka, A. and Harkins, J. 2001. ‘Introduction’, in J. Harkins and
A. Wierzbicka (eds), Emotions in Crosslinguistic Perspective, pp. 1–34.
Berlin: Mouton De Gruyter.
Wilson, R. 2013. ‘Another language is another soul’, Language and
Intercultural Communication 13 (3), 298–309.
Ye, V. Z. 2004. ‘La Double Vie de Veronica: reflections on my life as a Chinese
migrant in Australia’, Life Writing 1, 133–146.

23
English as a lingua franca
Ian MacKenzie
23.1 What is ELF?
English as a lingua franca (or ELF) is increasingly being used as an

alternative term for English as an international, global or world language
(although Canagarajah (2007) prefers ‘lingua franca English’, and Kecskes
(2007) and Mesthrie and Bhatt (2008) opt for ‘English Lingua Franca’).
Barbara Seidlhofer, one of the earliest researchers into ELF, defines it as
‘any use of English among speakers of different first languages for whom
English is the communicative medium of choice, and often the only
option’ (2011, p. 7). This definition does not exclude native English
speakers1 – for many of whom English is indeed the only option – but
most research into ELF focuses on bi- or plurilingual speakers using
English as an additional or auxiliary language.2
If English is used as a lingua franca, it is no longer related to a particular
native ‘target culture’ in which certain ways of speaking and behaving are
appropriate. Thus it is argued that rather than imitating the norms of
native English speakers, ELF users should adopt ways of speaking (with
their bi- or multilingual English-speaking interlocutors) which aid mutual
intelligibility and successful communication. And indeed corpus evidence
shows that spoken ELF contains a huge amount of linguistic variation and
non-standard forms, although formal written ELF (which generally allows
for more careful formulation and revision and re-reading) tends to resem-
ble English as a native language (ENL) to a much greater extent. To use
Chomsky’s (1986, p. 20) term, in ELF there is no stable E-language (external
language) as such; consequently ELF has been described by a panoply
of adjectives including adaptive, ad hoc, changing, contingent, creative,
diverse, dynamic, emergent, flexible, fluid, fragmented, fuzzy, heteroge-
neous, hybrid, indeterminate, mutable, open, shifting, unbounded,
unpredictable and unstable (e.g. Canagarajah 2007; Ferguson 2009; Firth
2009; Seidlhofer 2009b).

Multi-competence and English as a lingua franca 479
The variation to be found in ELF can be explained in different ways.

It may result from the fact that ELF users – ‘different constellations of
speakers of diverse individual Englishes in every single interaction’
(Meierkord 2004, p. 115) – are quite simply uninterested in the lexicogram-
matical norms of any particular native English speech community
(Seidlhofer 2001, 2009a, 2011). Alternatively, it might be a consequence
of the ‘shaky entrenchment’ (Mauranen 2012, p. 237) or the fuzzy proces-
sing (p. 41), or – from another perspective – the imperfect learning to be
expected in a second language (L2). Either way, the leitmotif of the
proponents of ELF is that it is different from, but not inferior to, ENL.
Cook (2002, p. 335) argued forcibly that L2 speakers need to be seen as
people in their own right – ‘the standards against which L2 users are
measured should be L2 user standards, not L1 native speaker standards.
Success should be measured by the ability to use the second language
effectively’ – and ELF researchers generally take a very similar line. As
Anna Mauranen (2006) puts it, we need to:
stop considering second and foreign language users as eternal ‘learners’
on an interminable journey toward perfection in a target language.
Speakers may opt out of the role of learner at any stage, and take on the
identity of language users, who successfully manage demanding
discourses despite imperfections in the code. (Mauranen 2006, p. 147)
Furthermore,
it is important for people to feel comfortable and appreciated when speak-
ing a foreign language. Speakers should feel they can express their iden-
tities and be themselves in L2 contexts without being marginalised on
account of features like foreign accents, lack of idiom, or culture-specific
communicative styles as long as they can negotiate and manage commu-
nicative situations successfully and fluently. (Mauranen 2003, p. 517)
Given that most ELF users are not emulating the idealised competence of
native English speakers, many orthodox SLA (second language acquisition)
and EFL (English as a foreign language) concepts are not relevant. ELF
researchers tend to write about divergent forms or features, rather than
errors, interlanguage, interference, fossilisation, etc. As Jennifer Jenkins
(2007) says, in international communication,
the ability to accommodate to interlocutors with other first languages
than one’s own (regardless of whether the result is an ‘error’ in ENL) is a
far more important skill than the ability to imitate the English of a native
speaker. (Jenkins 2007, p. 238)
Henry Widdowson (2003, pp. 48–49) describes ‘the nonconformist

usage’ of various adventurous authors (Carroll, Joyce, Cummings,
Achebe) as ‘evidence of the existence of . . . the virtual language, that
resource for making meaning immanent in the language which simply
has not hitherto been encoded’, and suggests that language learners (both

480 IAN MACKENZIE
L1 and L2) exploit this virtual resource too (thereby exercising what
Chomsky calls the creative aspect of language use). Seidlhofer (2011,
p. 120) describes ELF as ‘a different but not a deficient way of realising
the virtual language, or playing the English language game’: instead of
restricting themselves to the realisations of native English speakers, ELF
users exploit unused latent potentialities of English morphology, syntax
and phraseology.3
Dell Hymes (1972, p. 286) famously (at least among language teachers)
described ‘communicative competence’ in terms of ‘the systematically
possible, the feasible, and the appropriate’, but in relation to native
speaker norms; ELF speakers expand what is possible, feasible and
appropriate. As Widdowson puts it, the ELF perspective is that:
the modified forms of the language which are actually in use should be
recognised as a legitimate development of English as an international
means of communication. The functional range of the language is not
thereby restricted, but on the contrary enhanced, for it enables its users to
express themselves more freely without having to conform to norms
which represent the sociocultural identity of other people.
(Widdowson 2004, p. 361)
Of course, not everybody shares this ‘different but not deficient’ stand-
point on ELF; the orthodox EFL approach still expects non-native speakers
to imitate ENL, while conference interpreters tend to view ELF very
negatively, often preferring the acronym BSE, for ‘bad simple English’
(Reithofer 2010, p. 144).
Early work on ELF suggested that it might be codified and taught:
Seidlhofer (2001, p. 150) floated ‘the possibility of a codification of ELF
with a conceivable ultimate objective of making it a feasible, acceptable
and respected alternative to ENL in appropriate contexts of use’. Today,
however, Seidlhofer (2011, p. 77) argues that ELF should be ‘functionally
not formally defined; it is not a variety of English but a variable way of
using it’. The main focus of interest is the processes that underlie the
choice of particular linguistic features. Hence what is intrinsic to ELF –
what Alan Firth (2009, p. 150) calls the ‘lingua franca factor’ – is not any
specific language or discourse forms, but rather ‘the inherent interactional
and linguistic variability that lingua franca interactions entail’. Firth also
suggests that ELF users adopt a ‘lingua franca outlook’ on language, pre-
sumably as a result of their multilingualism – and their multi-competence.
Because they are by definition bi- or multilingual, EFL users (as opposed
to monolingual native English speakers using English in international
communication) are also likely to draw – consciously or unconsciously –
on their other language or languages. As Gibson Ferguson (2009) puts it,
ELF could be viewed:
as a fluid cluster of communicative practices where speakers draw on a
wide, not clearly bounded range of linguistic features – some standard,

some non-standard, and others not English at all (at least according to
the conventional view). (Ferguson 2009, p. 129)
Suresh Canagarajah (2007) compares the use of what he calls ‘lingua

franca English’ (LFE) with language use by radically multilingual commu-
nities in South Asia and Africa, where simultaneous childhood acquisition
makes it hard to say which language comes first, or to identify a mother
tongue or native language. This leads him to question whether languages
are actually ‘separated from each other, even at the most abstract level
of grammatical form’ (p. 923), and to suggest that ‘LFE raises serious
questions about the concept of language system’ (p. 926). Canagarajah
states that
LFE’s form is hybrid in nature. The language features words, grammatical
patterns, and discourse conventions from diverse languages and English
varieties that speakers bring to the interaction. Participants borrow from
each other freely and adopt the other’s language in their interaction with
that participant. (Canagarajah 2007, p. 926)
The type of multilingualism described by Canagarajah is in fact

relatively uncommon. For example, there are certainly fewer radically
multilingual communities in Europe than in South Asia and Africa, and
it is more common to talk about people coming from specific linguistic
backgrounds. Moreover, as Trudgill (1986) puts it,
The better-known European languages tend to be of the focused type: the
language is felt to be clearly distinct from other languages; its ‘boundaries’
are clearly delineated; and members of the speech community show a
high level of agreement as to what does and does not constitute ‘the
language’. (Trudgill 1986, pp. 85–86)
Many of the linguistic strategies attributed to LFE (or ELF) users, includ-
ing language mixing (or extensive code-switching and borrowing), are in
fact common to most bilinguals and multilinguals – or perhaps plurilin-
guals. The European Council for Cultural Co-operation’s Common European
Framework of Reference for Languages (2001, p. 4) uses ‘multilingualism’ to
refer to societies, or simply to ‘the knowledge of a number of languages’,
and ‘plurilingualism’ to refer to ‘a communicative competence to which
all knowledge and experience of language contributes and in which lan-
guages interrelate and interact’. However this distinction is not widely
shared, and the notion that multilingualism is somehow bad and plurilin-
gualism good is highly dubious. (Moreover, the word ‘plurilingual’ is
virtually unpronounceable for a great many Asian speakers of English.)
Yet the use of ‘linguistic features’ that are ‘not English at all’ seems to be
more common among European ELF speakers than, for example, in Asia.
This is because there are often typological similarities and/or cognate lexis
among neighbouring languages, which allows bi- and multilingual speak-
ers in Europe for whom English is a second or third language to exploit

482 IAN MACKENZIE
‘multi-competence’, or a dynamic multilingual system in which mental

representations from different languages interact, and more than one
language can be activated at the same time. Consequently, what might
be considered, in the use of a native language, as accidental transfer lapses,
erroneous words, deviant idioms and collocations, unconscious calques,
or even evidence of language attrition, can equally be seen as signs of
linguistic awareness and enhanced communicative competence when
used in ELF (examples will be given below).
23.2 ELF and multilingualism
François Grosjean (2010, p. 75) has long insisted on what he calls ‘the
bilingual or holistic view of bilingualism’, according to which ‘the bilin-
gual is an integrated whole who cannot easily be decomposed into two
separate parts’. Consequently, the bilingual ‘has a unique and specific
linguistic configuration. The coexistence and constant interaction of the
languages in the bilingual have produced a different but complete
language system.’ Cook (1991, p. 112) goes further, describing ‘the
compound state of a mind with two grammars’ as ‘multi-competence’ in
which the two languages are integrated, as opposed to coexisting. Other
ways of putting this are that – after a certain threshold is reached –
‘compound’ bilinguals have a conjoined system with a partially integrated
mental lexicon, or a ‘common underlying conceptual base’ (Kecskes and
Papp 2000).4
Grosjean (2008) further describes bilinguals’ states of activation of
languages and language processing mechanisms at any given time: ‘a
situational continuum ranging from a monolingual to a bilingual speech
mode’:
In the monolingual speech mode, the bilingual deactivates one language
(but never totally) and in the bilingual mode, the bilingual speaker
chooses a base language, activates the other language, and calls on it
from time to time in the form of code-switches and borrowings.
(Grosjean 2008, p. 38)
Grosjean insists (and most bilinguals would agree) that their languages are
never totally deactivated, and that there is constant interaction between or
among them: ‘bilinguals make dynamic interferences (ephemeral devia-
tions due to the influence of the other deactivated language) even in the
most monolingual of situations’ (p. 46).
Hence, given that ELF speakers are by definition bi- or multilingual, ELF
usage is likely to contain traces of dynamic interference – non-intentional
transfer or cross-linguistic influence5 from not-totally deactivated
languages – even if speakers try hard to remain in the monolingual
mode, as many Asian ELF users seem to do. Andy Kirkpatrick (2010)

discusses ‘code-mixing’ (the use of more than one language within the
same utterance, including borrowing, calquing idioms and creating hybrid
words) and the ‘use of lexical items and idioms with meanings specific to a
language variety’ and states categorically that:
One would not expect their use to be effective in lingua franca commu-
nication, as the specific lexical meanings would often be unknown to at
least some of the participants. Code-mixing would be an unlikely phenom-
enon in lingua franca communication because its use requires proficiency
in specific languages, and in lingua franca communication one could
never anticipate that all participants could possibly be familiar with all
the possible languages. After all, that is precisely why the participants in
lingua franca communication choose a lingua franca in the first place.
(Kirkpatrick 2010, p. 91)
This may be true in Asia, where there is often no overlap between lan-
guages. Just as bilinguals are not in the bilingual mode when speaking
with monolinguals, ELF speakers will not deliberately draw on languages
their interlocutors do not understand. Certainly the extracts from the
Asian Corpus of English (ACE) in Kirkpatrick (2007, 2010) show no sign
of borrowed words or neologisms, translated idioms, code-switching or
mixing, etc., in marked contrast to various national (nativised) varieties of
Asian English which notoriously mix in bits of other languages.6
However corpus data reveal that things are often different in ELF in
Europe. The major corpora available to researchers are the million-word
VOICE, the Vienna–Oxford International Corpus of English, consisting of
interviews, press conferences, service encounters, seminars, working
group and workshop discussions, meetings, panels, question/answer
sessions and conversations, and the million-word ELFA, the English as a
Lingua Franca in Academic Settings corpus, made up of 131 hours of
recorded lectures, presentations, seminars, thesis defences and confer-
ence discussions at the universities of Helsinki and Tampere in Finland.
In these corpora, code-switching – ‘a complete shift to the other language
for a word, a phrase or a sentence’ – and borrowing – ‘a morpheme, word
or short expression taken from the less activated language and adapted
morpho-syntactically (and sometimes phonologically) to the base
language’ (Grosjean 2008, p. 44) – are relatively common. Borrowing and
code-switching normally require the interlocutors to share the same
languages, but given the typological and lexical similarities among various
European languages, this is not necessarily the case with ELF. Theresa
Klimpfinger (2009, p. 348) states that ‘ELF, per definition, involves typically
three languages: the speakers’ first languages and English.’ Yet many
European ELF users speak (or at least understand) several languages, and
even if they don’t share their interlocutor’s L1, they may speak a typologi-
cally similar language. Many Europeans are capable of understanding a
large amount of cognate lexis, as a result of either multilingualism or

484 IAN MACKENZIE
implicit learning (or what Rita Franceschini (2009) calls ‘unfocused

acquisition’) – picking up (bits of) a language through direct contact with-
out any overt, conscious effort. Thus the particular nature of ELF, above all
in the linguistic environment of much of Europe (and among speakers of
more than one European language in Africa and South America), lends
itself to borrowing, code-switching and other forms of cross-linguistic
interaction. European ELF users can draw on their plurilinguistic
resources, and can sometimes operate in what Grosjean calls the ‘bilingual
language mode’ even if they do not share exactly the same language(s) as
all their interlocutors.
Despite being a Germanic language, English has so much Latin and
French lexis (as well, of course, as other loanwords from all around the
world) that it has been called a ‘semi-Romance language’ and (facetiously)
‘French badly pronounced’ (Barfield (1962, p. 59), quoted in Ó Laoire and
Singleton (2009), who also quote Pei’s (1967, p. 92) assertion that 12,000 of
the 20,000 words in ‘full use’ in English are of Latin, Greek and French
origin). Consequently, an English speaker has some familiarity with a large
number of the lemmas in ten national languages – French, Italian, Spanish,
Portuguese, Romanian, German, Dutch, Danish, Swedish and Norwegian –
as well as various other languages (e.g. Catalan) and dialects. It is hard for an
English speaker not to recognise a great many words in western European
languages, even if they are often spelled differently than in English (por
exemplo: arm, dishonesto, distinguir, estúpido, éxister, haº rd, histórico, illuminar,
katedral, Metall, natur, numeroso, parallèle, pedaal, populaer, razonable, recitare,
salade, seperato, Zement). This applies even more so to plurilingual European
ELF users, which often renders code-switching and borrowing viable
communicative strategies even when speakers do not share the same L1.
Clearly, most ELF speakers know less than linguists and language
teachers about language families, but a lot of experimental evidence
shows that language learners do have a strong sense of typological (or
psychotypological, or cross-linguistic) similarities and congruities among
languages (see Kellerman (1977) on ‘perceived language difference’ and De
Angelis (2007)). For example, Haº kan Ringbom (2007) shows that Swedish-
speaking Finns borrow Swedish rather than Finnish words when speaking
English, and Jasone Cenoz (2001) offers a similar demonstration of Basque
speakers borrowing from Spanish rather than Basque. Where languages are
similar, learners understand the L2 (or L3, etc.) more easily, and perceive
genuine similarities, which leads to positive L1 or ‘substrate’ transfer.
On the other hand, they are also often inclined to assume non-existent
similarities. It appears, as Kees de Bot (2004, p. 23) puts it, that ‘access to
words in the lexicon is non-selective’ so that ‘words from more than one
language compete for activation both in production and perception’. This can
result in unintentional lexical errors: there are thousands of false friends or
false cognates between English and European languages, most of which are
likely to lead to miscomprehension among people not sharing the same L1.

On the other hand, there are a small number of French words that have been
borrowed in most Germanic and Slavic languages (with a variety of spellings),
but which are used in (native) English with changed meanings. Well-known
examples include actual (meaning current or topical), eventual (possible),
possibility (opportunity), fabric (factory), concurrence (competition, competitors),
and sympathetic (friendly). All of these are false cognates in native English,
but would probably be comprehensible in context to the majority of
European ELF users. Consequently one might even argue that these are,
quite simply, true ELF cognates, or European English lexical items, which
native English speakers will need to reconceptualise as polysemous. From an
ELF perspective there seems little point in describing uses such as these as
‘negative transfer errors’ or ‘intrusions’, as one would from a mainstream
SLA perspective.
Although there are a small number of words that can be found in
most European languages, there are a great many others which are only
found in a few, usually in one language family. Some of these words
(along with grammatical features from particular language families)
may be borrowed or transferred into ELF. This can lead to ELF users
speaking recognisable varieties or ‘lects’ influenced by (and containing
recognisable transfer features from) a great many different and typolo-
gically diverse L1s.
Unlike regional dialects, which arise in local communities of speakers
talking to each other, these lects only come into prolonged contact with
each other in ELF, in linguistically heterogeneous situations. Because ‘they
arise in parallel, not in mutual interaction’, Mauranen (2012, p. 29) describes
these hybrid variants as ‘similects’, and says that ELF can be characterised as
‘second-order language contact’, a contact between hybrids:
Second-order contact means that instead of a typical contact situation
where speakers of two different languages use one of them in commu-
nication (‘first-order contact’), a large number of languages are each
in contact with English, and it is these contact varieties (similects) that
are, in turn, in contact with each other. Their special features, result-
ing from cross-linguistic transfer, come together much like dialects in
contact. (Mauranen 2012, p. 30)
Although Mauranen does not use the term, hybrid similects and second-
order language contact can easily be described as consequences of multi-
competence, or the knowledge of more than one language in the same
mind or the same community.
23.3 Lexis
Quite apart from nonce borrowings and more established loans from
speakers’ L1s, ELF corpora reveal a reasonable number of on-the-spot

486 IAN MACKENZIE
morphological or semantic approximations of established (but relatively

infrequent) English words, as well as one-off creations. Mauranen (2012,
p. 42) describes approximation as ‘involuntary and result[ing] from reali-
ties of perception, memory, and access’, and the ‘less deeply entrenched
memory representations’ involved in an L2 (p. 37). She suggests that taking
processing shortcuts and resorting to approximate forms of the language –
albeit forms which are reasonably close to target forms – is an efficient way
of using limited resources economically, because ‘the nature of processing
is fuzzy in most areas of cognition, including speech perception and
production’ (p. 41). But because ‘a complex environment like ELF seems
to require stretching the tolerance of fuzziness wider than usual’ (p. 42),
speakers seem to adapt to inexact words, making approximation a success-
ful communicative strategy. This endorsement of approximation
obviously contrasts with the orthodox SLA or EFL position, which is that
such errors should disappear as learners achieve greater familiarity with
the target language. This does not seem to be the case in ELF, in which
communication can succeed despite approximations.
Thus in the ELFA corpus, alongside many standard uses of irregular
past tenses, we also find a number of regularised ones, including binded,
bringed, digged, drawed, feeled, felled down, fighted, heared, losed, meaned, striked,
stucked and teached (forms which are also very commonly used by children
acquiring L1 English). ELFA also has not-quite-regularised past participles
including breaken, choosen, growned, sawn (for saw) and wonned. These
are used by speakers showing a high level of competence and fluency in
English, who sometimes also use the standard form in another speech-
turn. Another simplification strategy found in ELF appears to be ‘use the
negative prefix of your choice’: ELFA includes, among others, disbenefits,
discrease, inequal, inofficial, undirectly, unexperienced, unpossible, unrespect,
unuseful and unsecure.
Many other non-standard words found in ELF corpora result from
word formation processes that are widely attested in both native
English and the nativised New or World Englishes of Africa and Asia,
such as affixation, compounding, blending, conversion, modification
and shortening of the base (backformation and truncations) (see Plag
2003). Suffixation, conversion and modification, or what Mauranen
(2012, p. 126) calls the ‘extension of productive derivational principles
beyond their conventional boundaries’, can be seen in a large number
of words in ELFA, including the approximate or invented verbs
intersectioning, maximalise, plagiate, resoluting, satisfactionate, securiting and
successing; the nouns addictation, analytism, assaultment, assimilisation,
chemics, competensity, controversiality, governmentality, interventing, methodics,
militarians and paradigma; and the non-standard adjectives and adverbs
deliminated, devaluarised, disturbant, emperious, femininised, homogenic,
intentiously, methodologic, nationalisised, proletariatic, quitely, strategical
and theoretitised. There are also backformations, including colonisators,

introducted, presentate, registrate and standardisate, and truncations like

automously, categoration, decentralation, phenomen and significally. As
Mauranen says, ‘Morphology tends to be potentially overproductive in
natural languages. While convention and acquired preference keep it in
check in stable language communities, its possibilities are liberally
utilised by newcomers such as non-natives’ (2012, p. 127).7
There are also words that could come from a speaker’s L3 (or L4, etc.),
as well as their L1. For example, while phenomen was said by a speaker of
German (which has Phänomen), and homogene and prognose by a speaker
of Finnish (which has homogeeninin and prognoosi), instable and sportive
(athletic) used by German speakers, synthetise used by an Italian speaker,
and performant (efficient, competitive) used by a Romanian speaker, could
all be transfers from L3 French.
As well as form-based irregularities, ELFA shows many meaning-based
irregularities, when speakers use near-synonyms with a semantic affinity
to the more standard word, e.g. negated for denied (in they demand a power,
that is negated to them); strength for power and visioned for envisaged or seen (in it
was visioned as a wicked political strength); and normal for ordinary (in for normal
persons it’s really hard to understand). Mauranen points out that ‘in the
European context’, some of these approximations ‘have the additional
advantage of being related to a Latin origin, which would make the shared
semantic components familiar via other languages that the speakers
know’ (p. 103).
All of these creative approximations, or realisations of the virtual lan-
guage, have to be set against the underlying simplicity of ELF lexis.
Mauranen (2012, pp. 90–91) reveals that in the ELFA corpus, a mere 44
words (types) account for 50 per cent of the total words (tokens). These are
mainly monosyllabic, closed-class ‘grammatical’ words. The comparable
figure for MICASE, a corpus of spoken academic (American) English, is 58
words, from which Mauranen concludes that ‘ELF leads to notable lexical
simplification’ (p. 92). On the other hand, 38 of the most frequent 46 words
(i.e. 83 per cent) are shared by both ELFA and MICASE, showing that ‘when
it comes to fundamental vocabulary, first and second language use are
highly similar’ (p. 100).8
In the ELFA data, hearers occasionally use a standard English version of a
non-standard word in the next speech-turn (an embedded correction), but
corpus evidence suggests that faced with words like these, with what
Mauranen (2012, p. 103) calls ‘simultaneous deviance and recognisability
with respect to a standard form’, ELF users are more inclined to let them
pass, or indeed that they go unnoticed by both speakers and hearers. The
co-texts of non-standard words like those listed above show that ‘none
caused any noticeable reactions in their extended contexts’ (p. 103), or
seemed to provoke any ripples or breakdowns in communication.
In fact even when non-standard forms and linguistic anomalies do
(temporarily) hinder comprehension, Firth (1996, p. 243) suggests that

488 IAN MACKENZIE
ELF users often adopt the ‘let it pass’ principle: i.e. ‘when faced with
problems in understanding the speaker’s utterance’ – as long as the
problems are ‘non-fatal’ – the hearer ‘lets the unknown or unclear action,
word or utterance “pass” on the (common-sense) assumption that it will
either become clear or redundant as talk progresses’.
Let it pass is just one of a number of cooperative strategies described by
ELF researchers. Kirkpatrick (2010, pp. 127ff.), analysing data from south-
east Asian speakers, describes several more, including listening to the
message (i.e. disregarding non-standard forms); requesting repetition and
clarification when it is clear that a word is too important to let pass;
spelling out a word if pronunciation seems to be a problem; making things
explicit, including changes of topic; repeating a phrase (both self- and
other-repetition); speaker paraphrase or self-rephrasing (adjusting form
rather than meaning); participant prompting; participant paraphrase;
and lexical anticipation and lexical suggestion (offering a word if the
speaker hesitates or appears to stumble over a long word). Unlike the
multi-competent exploitation of cognates in European languages, interac-
tional strategies such as these can be used by all ELF speakers.
Convergent pragmatic strategies are often needed in ELF because of the
great range of accents, transfer features, and proficiency levels involved.9
Like speakers of World Englishes, ELF speakers often do not want or need
the competence stored in the mind of a native speaker; Shikaripur Sridhar
(1994, p. 802) calls this ‘the composite pragmatic model of bilingualism,
one that recognises that a bilingual acquires as much competence in the
two (or more) languages as is needed and that all of the languages together
serve the full range of communicative needs’. Yet limited competence in a
language can partly be compensated for by cooperative pragmatic strate-
gies, as well as by a tolerance of fuzziness and the exploitation of resources
created by the multi-competent integration of the mental lexicon.
23.4 Phraseology
The logic of approximation resulting from the absence of entrenched

schemas clearly applies to phraseology as much as to lexis. Corpora con-
firm that in ENL (and in many other languages – at least as used by their
native speakers) there is a phraseological tendency, which is to say that
units of meaning are largely phrasal, and speakers choose several words at
a time, often relying on memorised (entrenched, routinised, automatised,
conventionalised) chunks or formulaic sequences or prefabricated
phrases, rather than generating or improvising phrases ex nihilo (i.e. from
a vocabulary of thousands of lexical items and a stock of internalised
grammatical patterns or ‘rules’). Some of these phrases are entirely fixed,
others are semi-fixed, with slots that have to be filled according to context.
Native speech communities also have many, many thousands of strong

collocations, or words that habitually go together. Such phrases and

collocations are generally stored, produced and processed holistically,
rather than being analysed or decoded grammatically (see Pawley and
Syder 1983; Sinclair 1991, 2004; Wray 2002, 2008; Hoey 2005).
Yet a lot of evidence (and perhaps your own experience?) suggests that
classroom language learners are inclined to focus on grammatical form,
and to break down sequences, analyse their patterns, store small units,
and put them back together again when obliged to use the language.
As Alison Wray (2002, p. 206) puts it, ‘The result is that the classroom
learner homes in on individual words, and throws away all the really
important information, namely, what they occurred with.’ Breaking
down and analysing formulaic sequences and collocations (even if they
are recognised as such), in order to extract their lexical material, often
results in misremembering them and using them incorrectly (according to
native speaker norms).
Mauranen (2012, pp. 108ff.), however, contests Wray’s sweeping
claim that L2 learners fail to acquire stable collocations, or to speak in
memorised chunks, suggesting that evidence that ELF speakers do (attempt
to) process speech in chunks of language comes precisely from the
instances when the chunks go wrong, when speakers repeat whole
phrases when only one word needs to be changed: e.g. ‘I’ve been investigat-
ing this university so evaluating this universities’; ‘the point I said er earlier
that I made earlier’; ‘if we want to kind of have if we want to kind of set the
stage’. These rephrases clearly suggest that the processing unit is a chunk
rather than a single word, even if the desired chunk, not having been
encountered frequently enough, is insufficiently entrenched to be retrie-
vable at the first time of asking.
Mauranen (2012, p. 144) states that in L2 use, ‘speakers use sequences
which approximate conventionalised, phraseological units, but in ways
that do not quite match the target’, but – importantly – although ‘ELF
speakers tend to get them slightly wrong’, they ‘also get them approxi-
mately right’. In a frequent (if approximate) multi-word unit, the non-
standard elements ‘get embedded in the larger schematic unit, which
is sufficiently conventionalised to be recognisable as a unit in a given
function’. ELF corpora reveal lots of approximate forms of recognisable
ENL expressions, often using different articles and prepositions from
ENL use – e.g. by the time being, carved in stones, does it make any different,
from other point of view, in accordance to, in the right track, it’s very much
important, on the long-run, pieces by pieces, put the end on it, take closer look to
the world, there is quite much problems. Speakers also use semantic or
lexical approximations – e.g. building stones (for building blocks), divide
and govern (for divide and rule), the hen or the egg (for chicken and egg),
keep in the head (for bear in mind), lift an eyebrow (for raise), a side-product
(for by-product), small and middle business (for small and medium-sized
businesses), a streak of good luck (for stroke), turn a blank eye (for blind)

490 IAN MACKENZIE
(examples from Mauranen 2012, passim; Prodromou 2008, pp. 221–222;

and Pitzl 2012, p. 42).
Writing specifically about discourse markers that help organise interac-
tion, Mauranen (2009, p. 224) says that we might expect speakers ‘to use
them, because they are necessary for ensuring successful communication’,
and ‘to get them mostly right, because they are frequent and a restricted
class’. But she also expects them ‘to show variation in form that ignores
some conventions and preferences of native English speakers, because
English is a foreign language to most of those who use it as a lingua franca’.
ELF speakers also occasionally calque or literally translate idioms and
fixed expressions. This can be done between any languages, including
those that are not typologically related. Many writers on bilingualism
frown on calqued fixed expressions: e.g. Grosjean (2010, pp. 71–72) states
that ‘He was laughing in his fist’, translated from the German, ‘comes close
but isn’t quite right’ as ‘the correct English expression is “He was laughing
up his sleeve” ’. Unsurprisingly, the line usually taken in ELF research is
that calques are signs of idiomatic creativity rather than errors: in a hybrid
variety and speech community, there is no need to use standardised ENL
formulaic sequences, and every reason to calque or translate useful expres-
sions and idioms that you think will be understood. As Luke Prodromou
(2008) says,
when idioms do appear, they will appear in modified form, taking on the
shape of the mother tongue of the speaker and the pluralistic nature of the
speech encounter . . . ELF speakers will poach on L1 linguistic territory
when it suits them and when they are able to do so. Their use of phraseol-
ogy will be different from, but on an equal footing with, their L1-user
counterparts. (Prodromou 2008, pp. 251–252)
Calquing, or ‘poaching’ on L1 territory, can, of course, readily be rede-

scribed as a manifestation of multi-competence.
23.5 Grammar
The shaky entrenchment, fuzzy processing, approximation, creativity

and/or imperfect learning often found in ELF lexis and phraseology
obviously carries over to many grammatical constructions. Grammatical
accuracy, according to ENL norms, does not seem to be a priority. As
Mauranen (2012, p. 123) says, structural features ‘tolerate a good deal of
turbulence without disrupting communication’, and ‘ELF speakers show a
good deal of linguistic creativity in producing novel expressions that solve
the communicative problem at hand, even if they result in ungrammatical
forms or non-existent lexis from a Standard English viewpoint’
(pp. 199–200). Mauranen even concludes that ‘The variability of structure
in use and the frequency of non-standard structural features would appear

to call into question the significance of structure to successful commu-

nication’ (p. 130).
There would appear to be three major probable causes of the more
regular ‘grammatical features’ of ELF (or ‘ungrammatical forms’ according
to an orthodox EFL or SLA perspective). Some regularities involve
the simplification (and reconceptualisation) of inherently ‘difficult’ and
incongruous elements of English grammar – those which appear odd or
afunctional or idiosyncratic from the perspective of speakers of most other
languages. Other features in the grammar of specific ELF users are clearly
the result of cross-linguistic interaction with (or interference or transfer
from) particular L1s, as one would expect from a multi-competence
perspective. Still other features result from an orientation to increased
explicitness and clarity, which is often necessary in a heterogeneous and
unstable lingua franca.10 All such non-standard and non-native features
are in competition in ELF. Many of them probably go unnoticed a lot of
the time, but the simpler and more useful ones tend to be picked up and
re-used as speakers accommodate to each other’s uses.
An obvious afunctional oddity of Standard English is the third person
singular -s inflection (in an otherwise unmarked verb tense). Given the
accompanying third person pronoun, it is communicatively redundant,
and many ELF speakers sporadically use zero marking instead. Since
non-use of the third person -s has almost certainly not been taught, it
can be seen as a natural language development, in the expected direc-
tion of reduced redundancy, communicative economy and increased
regularity.
The use of prepositions is often variable in ELF, especially that of depen-
dent prepositions following verbs, nouns or adjectives (and preceding a
noun or gerund): ELFA includes accumulate to, focus to, an interest on, insisted
for, obsession in, satisfied at, etc., and Cogo and Dewey (2012, p. 57) cite
consequences on, depends of, implications on, and to influence on, as well as the
omission of prepositions, as in look this picture, listening music and it depends
the job. A contrary phenomenon is the creation of new prepositional verbs
by the addition of prepositions that give more emphasis, as in address to
(someone), answer to (a question), contact with, criticise about, discuss about,
emphasise on, mention about, phone to, reject against, return back, stress on, study
about, understand about, etc. (see Nesselhauf 2009; Seidlhofer 2011,
pp. 145ff.; Cogo and Dewey 2012, pp. 52ff.). Contrary to the zero third
person -s (and the regularisation of past tense forms, mentioned above),
these usages, which generally involve regularisation by analogy, compli-
cate rather than simplify the structure, and increase rather than reduce
redundancy (from an ENL perspective), but in the aid of clarity and
explicitness.
The use of articles in ELF often differs from ENL. Both a/an and the are
often used where ENL doesn’t use them (an advice, a progress, the nature, the
society), and omitted where ENL does have them (do lot of things, from Polish

492 IAN MACKENZIE
perspective, in Polish language), probably as a result of transfer from languages

that do not distinguish between countable and uncountable, and generic,
concrete and abstract nouns, and from languages without articles. Plurals
are frequently used with nouns that are uncountable in ENL. The English
categories of mass and count nouns, and the system of singulars and
plurals, are manifestly different from those of most other languages, and
speakers of many different L1s conceptualise nouns differently from
native English speakers, and consequently use nouns such as advice (or
rather, advices), equipments, evidences, feedbacks, furnitures and informations as
pluralisable count nouns.
Further easily recognisable grammatical features of ELF include the use
of the progressive rather than the present simple with stative verbs and
verbs with habitual meaning (e.g. I’m understanding, are belonging to, etc.);
the conflation of the present perfect and the past simple with definite past
time reference, so that there is no distinction between I have seen him
yesterday and I saw him yesterday; the use of the present simple rather than
the present perfect to express continuing time spans, as in He works here for
three years; a frequent blurring of the ENL semantic distinction between
since and for, with since being used for the duration of a time span (since three
years) as well as to refer to its starting point (since 2010); and verb comple-
mentation patterns which differ from those of ENL, with the to-infinitive
frequently used as a complement in place of the gerund.
Most of these uses are also widespread in the nativised World Englishes
of Africa and Asia (see Platt et al. 1984; Schmied 1991; Brutt-Griffler 2002;
Kortmann et al. 2004; Kachru and Smith 2008; Mesthrie and Bhatt 2008;
Trudgill and Hannah 2008). For speakers of many languages, they can be
seen as instances of L1 transfer, resulting in simplification, regularisation,
overgeneralisation, increased lexical and morphological transparency,
the loss (or exploitation) of redundancy, etc. Alternatively, as Erling
(2002, p. 11) suggests, rather than being examples of ENL having been
falsely analysed, perhaps some of these uses ‘are actually a symptom of a
change in the language which is coming about in non-native contexts but
is being held back by the standard English tradition’. Time will tell. But for
the moment, many of the particularities of ELF grammar can also be seen
as evidence of multi-competence in action.
23.6 Accommodation
The freedom for ELF speakers to vary patterns, lexis and morphology,
and use diverse and unpredictable non-standard forms deriving from
contact with many other languages, comes with a disadvantage. In ENL
(and most other L1s), hearers are always prospecting ahead on the basis
of the preceding context, rather than listening carefully to each word
(Sinclair 2004; Sinclair and Mauranen 2006). Words mostly come in

predictable patterns (Hunston and Francis 1999; Hoey 2005) so that ‘as
speech unfolds, hearers have usually already made guesses about what
is coming next, and need only confirm that their guesses were right’
(Mauranen 2012, p. 42). Such guesswork is more difficult in ELF, in
which patterns are more variable. However, lexicogrammatical variety
in ELF comes up against a countervailing force – people’s natural
tendency to accommodate to and converge with their interlocutors’
ways of speaking.
According to accommodation theory (see e.g. Giles et al. 1991), when
people interact they frequently adjust their speech (accent, dialect, lexis,
grammar, phraseology, etc.), so as to approximate the patterns of other
participants. Sometimes this involves borrowing or echoing interlocu-
tors’ exact words or phrases, and Mauranen (2012, p. 56) points out that
this is a good strategy in an L2, as these words ‘can be assumed to be
comprehensible to the speaker who first used them, and repeating them
(other-repetition or “echoing”) is one way of indicating comprehension’.
Peter Trudgill (2010, p. 189) states that accommodation is ‘a deeply auto-
matic process . . . the result of the fact that all human beings operate
linguistically according to a powerful and very general maxim’, which
Keller (1994, p. 100) phrases as ‘talk like the others talk’. Trudgill
continues:
Keller’s maxim, in turn, is the linguistic aspect of a much more general
and seemingly universal (and therefore presumably innate) human ten-
dency to ‘behavioural coordination’, ‘behavioural congruence’, ‘mutual
adaptation’ or ‘interactional synchrony’, as it is variously called in the
literature. This is an apparently biologically given drive to behave as one’s
peers do. (Trudgill 2010, p. 189)
In an L1, even the slightest variations of lexis, morphology, syntax,

phonology, prosody, etc. can be meaningful, as most language users are
aware of the social significance and/or geographical origin of speech var-
iants and lectal varieties. In ELF, much of this information is either lacking
or disregarded, and speech variation is less likely to carry class or geogra-
phical information. Instead, accommodation is more often related to
ensuring comprehensibility: ELF users – especially those who do not
know each other well, and cannot count on shared knowledge or linguistic
habits – adjust their speech in order to make it more intelligible to and
appropriate for specific interlocutors. As Edgar Schneider (2012) puts it,
ELF usage:
is a cybernetic process which involves the continuous monitoring
of the success of one’s contributions and, consequently, adaptive
behavior, i.e. linguistic negotiation in a broad sense. Hence, ELF
usage involves ongoing linguistic accommodation, a gradual increase
of shared communicative modes of performance in a specific setting.
(Schneider 2012, p. 60)

494 IAN MACKENZIE
Speakers accommodate to their addressees’ level of competence, simplify-

ing language use if necessary, and negotiate communicatively successful
forms. As we have seen, however, in ELF these forms may well differ from
those used in ENL.
23.7 ELF and multi-competence
Mauranen (2012, p. 32) believes that the natural tendency to linguistic

accommodation, combined with second-order language contact, will
lead to certain hybrid forms and innovations inspired by different L1s
spreading into common usage, so that ELF will become more stable.
The features likely to be diffused are those that are widely shared
among the world’s English speakers, because they are common in
other languages, and therefore acquired comparatively easily in
English as an additional language. As in all language contact, there is
also likely to be simplification, and a convergence or levelling of
phonological and grammatical systems, so that irregular and marked
features are gradually replaced by regular, unmarked alternatives. On
the other hand, given the heterogeneity of ELF speakers, variation will
remain, and the attempted codification and teaching of a particular
model of ELF would be both unwise – an attempt at ‘restriction by
partial incorporation’11 – and probably futile.
Even so, it is easy to overestimate the differences between ELF and
ENL. Both ELF and the nativised, postcolonial World English varieties in
Africa and Asia have largely appropriated the grammatical core of
the major native varieties of ENL, at least in their written forms, and
there is also a great deal of lexis common to all varieties. The claim that
native English speakers will need ‘an additionally acquired language
system’ in order to ‘communicate successfully in ELF settings where . . .
they are no longer the norm providers’ (Cogo and Jenkins 2010, p. 275)
is exaggerated – more interculturally competent speakers obviously do
adapt their speech, using the subconscious accommodation skills most
speakers develop; less interculturally competent speakers do not, or not
so much.12
All in all, ELF, particularly in Europe, seems to be a context in which
certain types of cross-linguistic transfer may indeed be acceptable and
effective. This does not furnish a global lingua franca – mixing western
European languages is unlikely to be an effective strategy with
Chinese, Iranian and Japanese ELF speakers – but it does facilitate a
situationally appropriate one. Rather than calling ELF a kind of per-
formance without competence, as Allan James (2000, p. 27) does, it
might be better to describe it as a widespread instantiation of multi-
competence.

Notes
1. Whatever may be argued elsewhere in this volume, I will not – except

here – put scare quotes around ‘native speaker’. The term has it uses;
as Trudgill (1995, p. 315) puts it pithily, ‘native-speaker’, like most
sociolinguistic concepts, ‘is not a matter of either-or’ but ‘a concept
which admits of degrees of more or less’, so that ‘some people are
more native speakers than others’.
2. As well as the native speakers of the inner circle (Kachru 1985, p. 12),
and the ELF users of the expanding circle, there are also the speakers of
nativised New or World Englishes of the outer circle. Consequently
Meierkord (2012, p. 2) suggests that ELF should be subsumed into a
broader category, ‘Interactions across Englishes’ (IaEs). She argues that
different Englishes potentially merge in these interactions, possibly
resulting in new emergent varieties, or ‘a heterogeneous array of new
linguistic systems’, but ones which do not exclude inner and outer
circle speakers.
3. This is a rather less elitist conception of potentiality than, say, that of
F. R. Leavis (1975), who insisted that ‘the full range of the English
language is there in its incomparable literature’ (p. 52); that ‘it is the
creative writer who maintains the life and potentiality of the lan-
guage’; and that ‘if mankind is to save, for future growth, its full
humanity’, we need to create ‘an educated public that, conscious of
its responsibility, shall maintain the “language” of creative thought
and keep the full potentiality alive’ (pp. 52, 242). Leavis is unlikely to
have approved of the potentiality inherent in Seidlhofer’s ‘virtual
language’; he notoriously accused Milton, whose poetry undeniably
reflects the (multi-competent?) influence of Latin constructions, of
having both ‘renounced’ and ‘forgotten the English language’ (1936,
pp. 52–53).
4. The term ‘compound bilinguals’ comes from Weinreich (1953/1968,
pp. 9–10). In compound bilingualism, concepts from two languages are
merged, unlike in coordinate bilingualism, in which translation
equivalents may be associated with different concepts, and subordi-
nate bilingualism, in which the conceptual representation of an L2
word is accessed via its L1 translation equivalent. Cook (2013, p. 35)
objects to the term ‘English as an additional language’, as for him it
implies that ‘becoming plurilingual is adding another language to
your first, not becoming a multi-competent speaker in whom both
are combined, maintaining in perpetuity Weinreich’s notion of
coordinate bilingualism’. However this interpretation of ‘additional
language’ is probably not shared by most people who use the term.
5. ‘Crosslinguistic influence’ was proposed by Kellerman and Sharwood
Smith (1986) as a neutral term, as opposed to the negative

496 IAN MACKENZIE
‘interference’. Herdina and Jessner (2002) prefer ‘crosslinguistic inter-

action’. There is as yet no accepted collective English term to cover
both willed and unwilled, positive and negative features of interaction
(borrowing, code-switching, transfer, interference, etc.), but in French,
Lüdi (1987, p. 2) has proposed marque transcodique.
6. For example many Singaporeans mix bits of Malay, Mandarin or other
Chinese dialects and Tamil into their English (widely known as
Singlish) when they are not using it as a lingua franca, while many
Filipinos mix English and Tagalog to the extent that the resulting
mixture has been named Taglish. However Kirkpatrick’s ACE data do
seem to show that Cogo’s (2009, p. 264) claim that ‘use of languages
other than English is so intrinsic to ELF as to be seen as a constituted
and constitutive part of it’ is Eurocentric and overstated.
7. While Mauranen prefers to describe lexical approximation as the
involuntary result of a lack of entrenched memory representations,
Pitzl et al. (2008) see only creative ‘innovations’.
8. Mauranen (2012, p. 91) further argues – convincingly – that language
contact and bilingual processing (of combined or competing systems)
in general tend to lead to lexical simplification, and the use of the most
frequent, and seemingly the most deeply entrenched, vocabulary.
This also occurs in learner language, and in translation, in which a
tendency to over-use core vocabulary and under-represent rarer words
has been posited as a (near) universal feature (Mauranen and Kujamäki
2004). This shows that it is not just imperfect learning that is at stake:
most translators have not learned their native languages imperfectly.
9. The range of accents in ELF is almost infinite. Jenkins (2000) has
proposed a Lingua Franca Core for Pronunciation, but this is not
uncontroversial; see MacKenzie (2014, Chapter 7), and the references
therein.
10. For a particularly thorough list of grammatical uses and non-uses one
should expect to find in ELF, see James (2000, p. 35).
11. This term originates with the political scientist Harold Lasswell, in
relation to strategies designed to defuse revolutionary ideologies or
cultural practices.
12. Moreover, it is rather unfair to assume that native English speakers are
necessarily incompetent monolinguals. After all, English speakers have
been in contact with speakers of other languages ever since the
first Angles, Saxons and Jutes settled in England. After the Viking
invasions, Old English and Old Norse speakers in contact probably
practiced receptive multilingualism, each speaking their own
language, as these were mutually intelligible with very similar pho-
nology, morphosyntax and lexis, including a lot of nearly identical
words, until eventually the Norse speakers shifted to English, possibly
simplifying the language in the process (McWhorter 2007, Chapter 4),
though it has also been argued that the great simplification of English

during the same period was due to Celtic speakers of Late British
slowly shifting to English, but somehow enticing the Anglo-Saxons
to abandon case endings and grammatical gender in the process
(Trudgill 2011, pp. 54ff.). After the Norman invasion there was a long
period in which both oral code-switching and written (triglossic)
language-mixing (in so-called macaronic texts) of French (or Anglo-
Norman), English and Latin seem to have been prevalent (Jefferson
and Putter 2013). All of this suggests that England has had more than
its fair share of multi-competent language users!
References
Barfield, O. 1962. History in English Words. London: Faber and Faber.

Canagarajah, A. S. 2007. ‘Lingua franca English, multilingual commu-
nities, and language acquisition’, Modern Language Journal 91 (5),
921–937.
Cenoz, J. 2001. ‘The effect of linguistic distance, L2 status and age on cross-
linguistic influence in L3 acquisition’, in J. Cenoz, B. Hufeisen and
U. Jessner (eds), Cross-linguistic Influence in Third Language Acquisition:
Psycholinguistic Perspectives, pp. 8–20. Clevedon: Multilingual Matters.
Chomsky, N. 1986. Knowledge of Language: Its Nature, Origin, and Use. New
York: Praeger.
Cogo, A. 2009. ‘Accommodating difference in ELF conversations: a study of
pragmatic strategies’, in A. Mauranen and E. Ranta (eds), pp. 254–273.
Cogo, A. and Dewey, M. 2012. Analysing English as a Lingua Franca: A Corpus-
Driven Investigation. London: Continuum.
Cogo, A. and Jenkins, J. 2010. ‘English as a lingua franca in Europe: a
mismatch between policy and practice’, European Journal of Language
Policy 2 (2), 271–294.
competence’, Second Language Research 7 (2), 103–117.
Cook, V. J. 2002. ‘Language teaching methodology and the L2 user perspec-
tive’, in V. J. Cook (ed.), Portraits of the L2 User, pp. 325–344. Clevedon:
Cook, V. J. 2013. ‘ELF: central or atypical form of second language acquisi-
tion?’, in D. Singleton, J. A. Fishman, L. Aronin and M. Ó Laoire (eds),
Current Multilingualism: A New Linguistic Dispensation, pp. 27–44. Berlin: De
Gruyter.
De Bot, K. 2004. ‘The multilingual lexicon: modeling selection and
control’, International Journal of Multilingualism 1 (1), 17–32.

498 IAN MACKENZIE
ELFA (English as a Lingua Franca in Academic Settings corpus).

www.helsinki.fi/elfa/elfacorpus
Erling, E. J. 2002. ‘ “I learn English since ten years”: the global English
debate and the German university classroom’, English Today 18 (2), 8–13.
European Council for Cultural Co-operation 2001. Common European
Framework of Reference for Languages: Learning, Teaching, Assessment.
Cambridge University Press.
Ferguson, G. 2009. ‘Issues in researching English as a lingua franca: a
conceptual enquiry’, International Journal of Applied Linguistics 19 (2),
117–135.
Firth, A. 1996. ‘The discursive accomplishment of normality: on “lingua
franca” English and conversation analysis’, Journal of Pragmatics 26 (2),
237–259.
Firth, A. 2009. ‘The lingua franca factor’, Intercultural Pragmatics 6 (2),
147–170.
Franceschini, R. 2009. ‘The genesis and development of research on
multilingualism: perspectives for future research’, in L. Aronin and
B. Hufeisen (eds), The Exploration of Multilingualism: Development of
Research on L3, Multilingualism and Multiple Language Acquisition, pp. 27–61.
Giles, H., Coupland, N. and Coupland, J. 1991. ‘Accommodation theory:
communication, context, and consequence’, in H. Giles, J. Coupland
and N. Coupland (eds), Contexts of Accommodation, pp. 1–68. Cambridge
University Press.
Grosjean, F. 2008. ‘The bilingual’s language modes’, in Studying Bilinguals,
pp. 37–66. Oxford University Press.
Grosjean, F. 2010. Bilingual: Life and Reality. Cambridge, MA: Harvard
University Press.
Hoey, M. 2005. Lexical Priming. London: Longman.
Hunston, S. and Francis, G. 1999. Pattern Grammar: A Corpus-Driven Approach
to the Lexical Grammar of English. Amsterdam: John Benjamins.
Hymes, D. 1972. ‘On communicative competence’, in J. B. Pride and
J. Holmes (eds), Sociolinguistics, pp. 269–293. Harmondsworth: Penguin.
James, A. R. 2000. ‘English as a European lingua franca: current realities
and existing dichotomies’, in J. Cenoz and U. Jessner (eds), English in
Europe: the Acquisition of a Third Language, pp. 22–38. Clevedon:
Jefferson, J. A. and Putter, A. (eds) 2013. Multilingualism in Medieval Britain
(c. 1066–1520): Sources and Analysis. Turnhout: Brepols.
Jenkins, J. 2000. The Phonology of English as an International Language. Oxford
University Press.
Jenkins, J. 2007. English as a Lingua Franca: Attitude and Identity. Oxford
University Press.

Kachru, B. B. 1985. ‘Standards, codification, and sociolinguistic

realism: the English language in the outer circle’, in R. Quirk and
H. G. Widdowson (eds), English in the World: Teaching and Learning the
Language and Literatures, pp. 11–30. Cambridge University Press.
Kachru, Y. and Smith, L. E. 2008. Cultures, Contexts, and World Englishes.
London: Routledge.
Kecskes, I. 2007. ‘Formulaic language in English Lingua Franca’, in
I. Kecskes and L. Horn (eds), Explorations in Pragmatics: Linguistic, Cognitive
and Intercultural Aspects, pp. 191–219. Berlin: Mouton De Gruyter.
Kecskes, I. and Papp, T. 2000. Foreign Language and Mother Tongue. Mahwah,
NJ: Lawrence Erlbaum.
Keller, R. 1994. On Language Change: The Invisible Hand in Language, translated
by B. Nerlich. London: Routledge.
Kellerman, E. 1977. ‘Towards a characterisation of the strategy of transfer
in second language learning’, Interlanguage Studies Bulletin 2, 58–145.
Second Language Acquisition. Oxford: Pergamon.
Kirkpatrick, A. 2007. World Englishes: Implications for International
Communication and English Language Teaching. Cambridge University Press.
Kirkpatrick, A. 2010. English as a Lingua Franca in ASEAN: A Multilingual Model.
Hong Kong University Press.
Klimpfinger, T. 2009. ‘ “She’s mixing the two languages together”: forms
and functions of code-switching in English as a lingua franca’, in
A. Mauranen and E. Ranta (eds), pp. 348–371.
Kortmann, B., Burridge, K., Mesthrie, R., Schneider, E. W. and Upton, C.
(eds) 2004. A Handbook of Varieties of English, vol. 2: Morphology and Syntax.
Berlin: Mouton De Gruyter.
Leavis, F. R. 1936. Revaluation: Tradition and Development in English Poetry.
London: Chatto and Windus.
Leavis, F. R. 1975. The Living Principle: ‘English’ as a Discipline of Thought.
London: Chatto and Windus
Lüdi, G. 1987. ‘Les marques transcodiques: regards nouveaux sur le
bilinguisme’, in G. Lüdi (ed.), Devenir Bilingue, Parler Bilingue, pp. 1–21.
Tübingen: Niemeyer.
MacKenzie, I. 2014. English as a Lingua Franca: Theorizing and Teaching English.
London: Routledge.
Mauranen, A. 2003. ‘The Corpus of English as Lingua Franca in Academic
Settings’, TESOL Quarterly 37 (3), 513–527.
Mauranen, A. 2006. ‘Signalling and preventing misunderstanding in
English as lingua franca communication’, International Journal of the
Mauranen, A. 2009. ‘Chunking in ELF: expressions for managing interac-
tion’, Journal of Intercultural Pragmatics 6 (2), 217–233.
Mauranen, A. 2012. Exploring ELF: Academic English Shaped by Non-Native
Speakers. Cambridge University Press.

500 IAN MACKENZIE
Mauranen, A. and Kujamäki, P. 2004. ‘Introduction’, in A. Mauranen and

P. Kujamäki (eds), Translation Universals: Do they Exist?, pp. 1–11.
Mauranen, A. and Ranta, E. (eds) 2009. English as a Lingua Franca: Studies and
Findings. Newcastle upon Tyne: Cambridge Scholars.
McWhorter, J. 2007. Language Interrupted: Signs of Non-Native Acquisition in
Standard Language Grammars. New York: Oxford University Press.
Meierkord, C. 2004. ‘Syntactic variation in interactions across interna-
tional Englishes’, English World-Wide 25 (1), 109–132.
Meierkord, C. 2012. Interactions across Englishes: Linguistic Choices in Local and
International Contact Situations. Cambridge University Press.
Mesthrie, R. and Bhatt, R. M. 2008. World Englishes: The Study of New Linguistic
Varieties. Cambridge University Press.
Nesselhauf, N. 2009. ‘Co-selection phenomena across New Englishes:
parallels (and differences) to foreign learner varieties’, English World-
Wide 30 (1), 1–26.
Ó Laoire, M. and Singleton, D. 2009. ‘The role of prior knowledge in L3
learning and use’, in L. Aronin and B. Hufeisen (eds), The Exploration of
Multilingualism: Development of Research on L3, Multilingualism and Multiple
Language Acquisition, pp. 79–102. Amsterdam: John Benjamins.
Pawley, A. and Syder, F. H. 1983. ‘Two puzzles for linguistic theory: native-
like selection and nativelike fluency’, in J. C. Richards and R. W. Schmidt
(eds), Language and Communication, pp. 191–225. London: Longman.
Pei, Mario 1967. The Story of the English Language. London: George Allen and
Unwin.
Pitzl, M.-L. 2012. ‘Creativity meets convention: idiom variation and
re-metaphorization in ELF’, Journal of English as a Lingua Franca 1 (1),
27–55.
Pitzl, M.-L., Breiteneder, A. and Klimpfinger, T. 2008. ‘A world of words:
processes of lexical innovation in VOICE’, Vienna English Working Papers
17 (2), 21–46 (online).
Plag, I. 2003. Word-Formation in English. Cambridge University Press.
Platt, J., Weber, H. and Ho, M. L. 1984. The New Englishes. London: Routledge
and Kegan Paul.
Prodromou, L. 2008. English as a Lingua Franca: A Corpus-Based Analysis.
London: Continuum.
Reithofer, K. 2010. ‘English as a lingua franca vs. interpreting: battle-
ground or peaceful coexistence?’ The Interpreters’ Newsletter 15, 143–157.
http://hdl.handle.net/10077/4755
Ringbom, H. 2007. Cross-linguistic Similarity in Foreign Language Learning.
Schmied, J. 1991. English in Africa: An Introduction. London: Longman.
Schneider, E. W. 2012. ‘Exploring the interface between World Englishes
and second language acquisition – and implications for English as a
lingua franca’, Journal of English as a Lingua Franca 1 (1), 57–91.

Seidlhofer, B. 2001. ‘Closing a conceptual gap: the case for a description of

English as a lingua franca’, International Journal of Applied Linguistics 11 (2),
133–158.
Seidlhofer, B. 2009a. ‘Accommodation and the idiom principle in English
as a lingua franca’, Journal of Intercultural Pragmatics 6 (2), 195–215.
Seidlhofer, B. 2009b. ‘ELF findings: form and function’, in A. Mauranen and
E. Ranta (eds), pp. 37–59.
Seidlhofer, B. 2011. Understanding English as a Lingua Franca. Oxford
University Press.
Sinclair, J. 1991. Corpus, Concordance, Collocation. Oxford University Press.
Sinclair, J. 2004. Trust the Text: Language, Corpus and Discourse. London:
Routledge.
Sinclair, J. and Mauranen, A. 2006. Linear Unit Grammar. Amsterdam: John
Benjamins.
Sridhar, S. N. 1994. ‘A reality check for SLA theories’, TESOL Quarterly 28 (4),
800–805.
Trudgill, P. 1986. Dialects in Contact. Oxford: Blackwell.
Trudgill, P. 1995, ‘Linguistic oppression and the non-native speaker’,
Trudgill, P. 2010. Investigations in Sociohistorical Linguistics: Stories of
Colonisation and Contact. Cambridge University Press.
Trudgill, P. 2011. Sociolinguistic Typology: Social Determinants of Language
Complexity. Oxford University Press.
Trudgill, P. and Hannah, J. 2008. International English: A Guide to the Varieties of
Standard English, fifth edition. London: Edward Arnold.
VOICE (Vienna–Oxford International Corpus of English). www.univie.ac
.at/voice
Weinreich, U. 1953/1968. Languages in Contact: Findings and Problems. The
Hague: Mouton.
Widdowson, H. G. 2003. Defining Issues in English Language Teaching. Oxford
University Press.
Widdowson, H. G. 2004. ‘A perspective on recent trends’, in A. P. R. Howatt
and H. G. Widdowson, A History of English Language Teaching, second
edition, pp. 353–372. Oxford University Press.
Wray, A. 2002. Formulaic Language and the Lexicon. Cambridge University
Press.
Wray, A. 2008. Formulaic Language: Pushing the Boundaries. Oxford University
Press.

24
A critical reaction from
second language
acquisition research
David Singleton
24.1 Introductory
This chapter is billed as a critique of multi-competence from an SLA point

of view. Actually, although I happily accept the role of riding under the SLA
banner – as does Vivian Cook still on occasion (see Cook and Singleton
2014) – I am not an SLA die-hard, as my many (and enthusiastic) incursions
into the world of multilingualism studies demonstrate (e.g. Aronin
and Singleton 2012; Singleton et al. 2013; Pfenninger and Singleton,
forthcoming). Nor is my perspective on multi-competence especially cri-
tical – certainly no more so, in general, than that of a number of the
contributions to this volume (including Cook’s) which speak of the infancy
of the concept and of its uncertain and speculative nature. I heartily
subscribe to the basic perspective (Cook, Chapter 1, this volume) which
speaks of there being an overarching, overall system discoverable in any
mind or community that uses more than one language variety. There is,
however, one dimension of the rhetoric of multi-competence which wor-
ries and dismays me in terms of the way it can be interpreted, and is
sometimes clearly intended to be interpreted. My worry and dismay in
this connection may well derive from my SLA conditioning, but, in any
case, I wish to make again, with renewed vigour, some of the points I have
made before on this matter elsewhere (see e.g. Singleton 2012).
The point at issue is that there seems to be a temptation, often a
positive desire, amongst some multi-competence enthusiasts – but also,
let it be said, more widely – to adopt a line of discourse which can be
read as positing a degree of unitariness of language knowledge that
seems very much at odds with how language knowledge has usually
I should like to thank Simone Pfenninger for her very insightful comments on an earlier version of this text.

A critical reaction from second language acquisition research 503
been seen as operating. Such discourse is, it has to be said, in many cases
just a way of putting a gloss on the proposition which I fully accepted in
the last paragraph. Thus, when Scott (Chapter 21, this volume) refers to
foreign language learning as an integral part of the ecosystem and Seton
and Schmid (Chapter 16) use the term supersystem, they are really just
referring to the notion of overarching system discussed earlier.
Similarly, when Aronin (Chapter 7) states that her Dominant Language
Constellation can be construed as one language, this again in the given
context means something similar to ‘overall system’, as the rest of her
chapter makes clear. Sometimes, when researchers make claims about
the overall effect of multilingualism on, for example, cognition
(Athanasopoulos, Chapter 17), personality (Dewaele, Chapter 19), or
creative abilities (Kharkhurin, Chapter 20), a radically confluent out-
come of the interaction between the language varieties learned/known
is indeed suggested, but not necessarily a radical confluence of the
knowledge associated with each of the language varieties in question.
(There is certainly no suggestion of this kind of confluence in the writ-
ings of the authors mentioned.) Of course, many researchers do make
reference to influence and compromises between the knowledge sys-
tems relating to individual language varieties (Hirata-Edds and Peter,
Chapter 15; Ortega, Chapter 3; Seton and Schmid, Chapter 16; Su,
Chapter 14), but the usual view (and the view of the persons referred
to) is that multi-competent versions of individual language systems do
not completely blend in the mind, even in regard to particular features.
There is, nevertheless, a tendency on the part of some writers to appear
to want to abandon or abolish all talk whatsoever of boundaries between
language varieties in the mind, a tendency which may perhaps find com-
fort in usage-based accounts of language knowledge as dynamic, provi-
sional and malleable semiotic constellations (Hall, Chapter 9). This
abandonment of boundaries is quite a standard view in respect of the
semantic or conceptual dimensions of language (MacKenzie, Chapter 23),
where the received wisdom is that there is a ‘common underlying con-
ceptual base’ for all language varieties known by the multi-competent user
(Kecskes and Papp 2000). The desire to dispense with boundaries does not,
however, stop there; increasingly researchers – of a range of theoretical
stances – are tending to call into question the notion that languages in the
mind in general are clearly bounded entities (e.g. de Bot, Chapter 6, this
volume; de Bot, Lowie and Verspoor 2007; Franceschini, Chapter 5; Harris
1998; Toolan 2008; Vaid and Meuter, Chapter 4). My response to this
position is that language knowledge in the mind in all its aspects is, in
fact, highly differentiated, and that such differentiation broadly follows
the lines recognised by the traditional boundaries which draw (always, of
course, crossable and permeable) lines between languages. I would assert
that it is only by recognising such lines of demarcation, such differentia-
tion and differentiability, that we can explain a very large number of

504 DAV I D SI N GL ETO N
linguistic phenomena, including everyday bilingual and multilingual

communication.
24.1.1 Unbounded?
Multi-competence has, as one would expect, been used as a watchword by
those whose perspective emphasises the close relationship between the
various kinds of language knowledge and of cultural knowledge available
to an individual functioning in more than one language and culture. For
instance, Kanno, writing about the negotiation of bilingual/bicultural
identity, has the following to say:
I take the ‘multi-competence’ (Cook 1992, 1999) view of bilingualism and
biculturalism: that is, to think of a bilingual person’s linguistic and
cultural repertoire as a whole, rather than separating out each language
and culture as if ‘the bilingual is (or should be) two monolinguals in one
person’ (Grosjean 1992: 52). (Kanno 2003, p. 7)
This kind of statement is not uncommon, as is evident from the discussion

in the last section.
Of course, no one would wish to deny the close connectivity characteris-
ing the relationship between diverse aspects of the languages and
language varieties that the bilingual or multilingual uses. Long gone are
the days when anyone has seriously wished to advance the notion of such
an individual being, as it were, two or more monolinguals in one person.
One can cite in this connection Safont Jordà’s comments on the interde-
pendence of the trilingual’s language knowledge systems:
This interdependence characterising third language learning leads us to
consider learners’ first second and third languages as a whole linguistic
system, which they command simultaneously. In fact, it seems more
logical to consider languages known by a multilingual speaker as a
whole unit than to view them as separate entities that develop in isolation.
(Safont Jordà 2005, pp. 13–14)
There is still a question-mark, however, over just how far we should take
such integrationist arguments. The available empirical evidence does not,
I would argue, amount to any kind of convincing case for the idea of an
undifferentiated mass; in fact, it points in precisely the opposite direction.
On this matter Cook seems to have wavered a little, to have succumbed
to ambivalence. His earlier position (2003, p. 7f.) on the relationship
between two languages known to an individual was that while ‘total
separation is impossible since both languages are in the same mind’, it is
also the case that ‘total integration is impossible since L2 users can keep
the languages apart’. He went on to suggest that ‘between these two
extreme, and probably untenable positions of total separation and
total integration, there are many different degrees and types of

interconnection’. In his latest pronouncements on this topic (e.g. this

volume) he appears to take a somewhat different line; he says that whilst
it is easy to slip into suggesting that there are objects called languages in
the mind which relate together like the coloured balls in Bohr models of
the atom, on another understanding of language the phenomenon
resembles more an amorphous, indeterminate entity from quantum
theory than a set of clearly defined objects. This looks like a proposition
very close to de Bot’s speculation in this volume (which he, interestingly,
opposes to the multi-competence notion), that language in the mind may
be one merged system of situation-specific utterances.
It is difficult to know sometimes whether the dramatic images Cook
deploys – for example, the likening of language in the mind to an
amorphous entity from quantum theory – are just manières de parler.
After all, his first premise regarding multi-competence (Chapter 1, this
volume) is that multi-competence concerns the total system for all
languages (L1, L2, Ln) in a single mind or community and their inter-
relationships. In relation to what goes on in the mind, he says that
languages must be an inter-connected whole within a single mind, an
ecosystem of mutual interdependence. This resembles Safont Jordà’s
suggestion, and is entirely unproblematic from my point of view, since
inter-relationship, inter-connection and interdependence imply, not
amorphousness, but a multi-dimensional structure, a plurality of
elements, which are linked to each other and dependent on each
other, but differentiable.
24.2 Some evidence
A powerful argument against the unboundedness of language knowledge

in the minds of bilinguals and multilinguals derives from the fact of formal
differences among language varieties. Thus, faced with the challenge of
coming to grips with the formal shape of an unfamiliar word, a person will
have recourse to consulting the formal structure of more familiar items
and will then analogise (see Bybee 1988, Stemberger and MacWhinney
1988; cf. Gentner 2010, Huang and Pinker 2010). Now, the languages
known to such an individual may have highly divergent morphological
phonological systems. Take the case of a Chinese–English bilingual who
comes across the word explicitness for the first time. Such a person will refer
to the structure of words he/she does know, like happiness and sadness.
Given that /nəs/ does not figure as a morpheme or a syllable in Chinese,
it is highly probable that the search on which such analogising depends
has to run through the lexicon of each language separately. In other words,
concrete formal differences between languages have clear implications for
the way in which they are processed, and are thus necessarily attended by
differentiation in the mind.

Another kind of evidence of boundedness comes from the phenomenon

of the selective recovery of language varieties – including language vari-
eties popularly termed dialects – lost as a result of brain damage. There is
also some mysterious but significant evidence from non-parallel aphasia
(see, e.g., Fabbro 1999, Chapters 12–16). For example, Whitaker (1978,
p. 27) refers to the case of an English scholar in the area of the classical
languages and literatures who recovered first Ancient Greek, then Latin.
He subsequently recovered French and finally English (his L1). Grosjean
(1982, p. 260), for his part, refers to the instance of a native speaker of
Swiss German (which is, of course, very different from Standard German)
who recovered first French and later Standard High German, but who
never recovered his native variety. Fabbro (2002, p. 204) reports the strange
and fascinating case of an Italian native speaker of Veronese (a variant of
Venetian, very different from Standard Italian), who had exclusively used
Veronese in all her daily activities, except for a few words of Italian two or
three times a year, but who, following a brain insult, communicated
exclusively in Standard Italian. Her condition subsequently improved to
the point where she could understand Veronese, but she persisted in
producing only Standard Italian. A still more recent case is that of Jürg
Schwyter (Schwyter, 2011), who, following a stroke, lost the use of every
one of his languages. He recovered receptive capacities in all his languages
Swiss German, Standard German, English, Italian and French, but recov-
ered full productive capacities only in his mother tongue, Swiss German,
and his main professional language, English.
With respect to non-parallel aphasia, Paradis and Goldblum (1989)
report the case of a trilingual subject who was a native speaker of
Gujarati, an Indo-European language spoken in central India, which was
the language he mainly used with his family. However, the person in
question lived in Madagascar, and so he had additionally acquired
Malagasy, Madagascar’s official language. At age six he had also learned
French at school, and he used this language on a daily basis in his profes-
sional activities. Following a neurosurgical operation, he evidenced
disorders typifying Broca’s aphasia in Gujarati but no deficits in his other
languages. Two years after the operation he had fully recovered Gujarati
but had difficulties with Malagasy in terms of verbal fluency and syntactic
comprehension. Four years after the operation no disorder was detected in
either language. A not dissimilar case is reported by Gil and Goral (2004).
This involved a 57-year-old Russian–Hebrew bilingual, who was initially
diagnosed with expressive–receptive aphasia in both languages,
subsequently with predominantly receptive aphasia in both languages.
Then, three-and-a-half months post-onset he was diagnosed differently in
the two languages: he was deemed to be suffering from predominantly
receptive aphasia in Hebrew and from amnestic aphasia (word-finding and
object-naming difficulties) in Russian. Six weeks later he was diagnosed as
amnesiac in both languages.

These cases of selective recovery and non-parallel aphasia, in my view,

bespeak boundedness. Note that this term is not used here in a strong
sense of disconnection; even languages which are never recovered by an
individual are undoubtedly still at some level ‘there’ and connected to
others, even though these latter may not be available. Nor is it intended as
an implicit claim that languages are sited in different physical locations in
the brain (see e.g. Fabbro 2001). An explanation of language loss of this
kind which is more in harmony with the neuroimaging evidence is that it
is due to damage to control mechanisms located in the prefrontal cortex
that activate the target language and inhibit the non-target language
(Pfenninger, personal communication). Boundedness in the present
context betokens, more weakly, that each language in the mind has its
own processing dynamic – has a degree of, as it were, autonomy, of
developmental individuality.
On the early developmental front, the individuality of the progress of
each language is indicated by studies such as those conducted by
Schelleter, Sinka and Garman (e.g. Schelleter, Sinka and Garman 1997;
Sinka and Schelleter 1998; Sinka, Garman and Schelleter 2000). The
studies in question looked at two children acquiring, respectively,
Latvian and German (both highly inflected languages) alongside English
(not a very inflected language). They provided evidence of the develop-
ment of functional categories in Latvian and German from the earliest
stages, but not in English, from which the researchers conclude that the
nature of Latvian and German input is rich enough to trigger functional
category development, whereas the English input is not.
The strong implication of these last findings is that the languages
acquired by a simultaneous multilingual develop separately. The question,
however, of whether this is in fact the case from the very earliest stages of
acquisition has been quite a controversial one. A highly influential
perspective on this matter was that the simultaneous multilingual began
with a single language system and that his/her languages separated only at a
later stage (e.g. Volterra and Taeschner 1978; see discussion in Clark 2009,
p. 342ff.). This hypothesis suggests that simultaneous multilinguals begin
with a single language system, a single fused linguistic representation, and
that it is only around the age of three years that they begin to differentiate
their languages (see e.g. Pettito et al. 2001, p. 455). According to this view,
the child in the very early stages of language development is not in posses-
sion of translation equivalents across languages, but rather he/she has a
single lexical store, with a single word from one or other of his/her lan-
guages for any given meaning. On this basis, the evidence cited in favour of
the above optique tended to be that of language mixing (cf. Macrory 2006,
p. 163). The claim was that mixed utterances arose because the child at an
early stage did not have access to translation equivalents across languages,
that he/she had just one lexical store, with a particular word from one or
other of his/her languages for any given meaning.

It is irrefutable that language mixing goes on in the language use of

young multilinguals, and that much of this mixing happens because a
child may know an expression in one language for which he/she has no
equivalent in other languages. Nicoladis and Secco (2000), for example,
report that around 90% of this kind of mixing that they observed in very
young bilinguals was explicable in terms of lexical gaps in one language
or the other. That is to say, when the children lacked the word they
needed in one language, but had it at their disposal in their other
language, they simply drew on what they knew to supplement what
they did not know. This strategy undoubtedly continues throughout
childhood and indeed into adulthood. In an MPhil project on Chinese–
English development, Zhang (2006) demonstrates this with respect to
sibling–sibling interaction between two Chinese–English bilingual
children, where, for instance, the Chinese expressions kao-ya (‘roast
duck’) and fu lu (‘pickle made from soya beans’) were used in English
matrix utterances because the English translation equivalents were
unknown (and in the latter case non-existent). This is a very natural
strategy for the multilingual child to adopt. It of itself says nothing
about the question of the separation or integration of a young multi-
lingual’s languages. Quay (1995), for her part, shows the falsity of the
notion that the multilingual’s lexicon is systematically distributed
across languages; and she, accordingly, strongly disputes the claim
that there is a stage at which the multilingual has just one item in
one or other language for a particular meaning. In her own case study
she traced the lexical development of an infant acquiring Spanish and
English from birth to the age of 1;10 via daily diary records and weekly
video recordings in the two language contexts. What emerged from this
study was that translation-equivalents that make possible language
choice were available from the beginning of speech. In a later study
Deuchar and Quay (2000) demonstrated that differentiation between the
young bilingual’s two language systems is discernible also in phonology
and morphosyntax.
Bilingual/multilingual children, in other words, generally keep their
languages apart when using them, and they are highly adept – even at a
very early age (e.g. Genesee, Nicoladis and Paradis 1995; Nicoladis 1998) –
at making decisions as to which language to speak to whom. It seems,
moreover, that on occasions where languages are mixed, the mixing in
question may evidence an awareness – again from an early age – of
differentness of their languages (cf. Meisel 2008) and of the different
language competencies of interlocutors (see e.g. Lanza 1997). According
to one theoretical position, the mind is genetically programmed to impose
structure on linguistic material in developing grammars (Berkes and
Flynn, Chapter 10, this volume). Whatever the truth of this proposition,
the mind is certainly encouraged to impose differentiating structure by
communicative necessity (see below). De Houwer puts it this way:

Like monolingual children, bilingual children pay a lot of attention to the

input they receive. They soon notice that this input differs depending on
who is talking and in what situation someone is talking. Just like mono-
lingual children, bilingual children attempt to talk like the people around
them. Because of the bilingual situation, however, the bilingual child has
more options than the monolingual one . . . [A]t a very young age bilingual
children are skilled conversationalists who easily switch languages.
(de Houwer 1995, p. 248)
A further kind of language loss also points to the boundedness of

languages. Languages may be lost not only as a result of stroke and other
kinds of brain damage but also as a result of neglect. Where, for example, a
child migrating to a particular community has an L1 which is different
from that of the majority of members of the host community, where his/
her home language receives little or no support from the community in
question and where his/her parents make little or no effort to support it
either, the language in question may attrite to the point of virtual disap-
pearance. Case-studies of such L1 attrition are reported by Kouritzin (1999).
One such instance (Kouritzin 1999, pp. 75–96) is that of Lara, who had
migrated with her family from Finland to Canada at age two, and had then
lived for four years in a small town within a tight-knit Finnish community.
Lara was thus until age six a Finnish speaker with very little English. From
then onwards, however, having moved to a large city, and under the
influence of her parents’ decision that the time had come to integrate
with English-speaking Canada, her development in Finnish came to a halt
and English progressively took over. By her own account, the last time she
tried (and failed) to converse in Finnish had been when she was eighteen
years old. Her perception is that she has lost her L1. Here we have an
example of a family taking a decision to favour one language and effec-
tively to abandon another, with the result that the latter language was lost.
This clearly implies choice and the boundedness of different languages.
The opposite kind of choice is made in the case of the decision of an
individual to learn or revive knowledge of a particular language. For
example, an individual may opt to identify with a particular heritage
language (see Hirata-Edds and Peter, Chapter 15, this volume) and attempt
to acquire it or to revitalise his/her command of it. Again the possibility of
the kind of determined and specific language-allegiance involved in this
kind of choice bespeaks the boundedness of the languages in question.
To conclude this section, it is worth referring to the fact that in
their normal functioning multilinguals are very attentive in their use
or non-use of specific languages to the competences of their interlocu-
tors. This is clearly a necessary condition of successful communication.
The bounds of a language in the language user’s mind are thus set by
the limits of intelligibility. Of course there are plenty of cases of
multilingual societies – e.g. Singapore – where there is a promiscuous
intermingling of languages. In societies like Singapore, however, it

should be noted that the intermingled languages in question are acces-

sible to the community at large. Sensitivity to which languages will and
will not work in a given context obviously is an everyday given in the
life of a multilingual. It is interesting to explore in this context Li Wei’s
(2011) discussion of the notion of translanguaging (see also Garcia and Li
Wei 2014; Kharkhurin, Chapter 20, this volume; Li Wei, Chapter 8),
which he characterises as both operating between different linguistic
structures and systems and going beyond them. He also sees trans-
languaging as covering the full range of linguistic performances of
multilingual language users. He represents the act of translanguaging
as transformative, creating a social space for the multilingual by bring-
ing together different dimensions of experience, attitudes and
cognitive capacity into one coordinated and meaningful performance.
What is patently not being talked about here is language confusion, or
an absence of language differentiation. On the contrary, the focus is on
opting between language resources, on the coordination of such resources
and on multilinguals’ constant decision-making as to what strategic
moves they should make to achieve specific communicative effects.
Code-switching is certainly not counter-evidence to the notion of
language boundedness. It can without a doubt be seen (de Bot and
Schreuder 1993) as evidence favouring the notion that a multilingual’s
various languages are continuously activated, though each to a different
level (see Paradis 1981; Green 1986), and thus allowing for the possibility
that words from the language other than the one in use may occasionally
‘slip in’ (de Bot and Schreuder 1993, p. 212), but this does not undermine
the essential boundedness of a multilingual’s languages. One might add
that the manner in which code-switching proceeds appears to be sensitive
to the particularities of the languages involved. For example, Myers-
Scotton (2003) notes that code-switching between Arabic and English
exhibits a very high proportion of embedded English inflectional phrases
(i.e. phrases including a tensed verb) where Arabic is the matrix language.
Myers-Scotton argues that the reason for this untypically large number of
embedded language ‘islands’ is the essential incongruity between the
Arabic frame and the nature of English verbs. Such sensitivity to cross-
linguistic incongruities constitutes further evidence of the boundedness of
languages in the mind.
The above arguments and evidence seem to me cumulatively to make a
strong general case for retaining a notion of boundaries between
languages in the mind. We move now to examine three very particular
issues, touched on already, in relation to which the debate about bound-
edness is perhaps at its sharpest: the question of whether the knowledge
associated with a particular language is really activated as a whole set; the
question of whether at the conceptual level languages are really fused in
the mind; and the question of how to view the deployment of what looks
like an undifferentiated mish-mash of language knowledge.

24.3 Three issues
24.3.1 The activation of languages as bounded sets of knowledge

In relation to the issue of whether the knowledge associated with a parti-
cular language is activated as an individual whole set, a good reference
point is the work of Dijkstra and his colleagues (e.g. Dijkstra 2003; see also
de Bot 2004). Dijkstra’s perspective rests on findings from a range of
experimental studies (including his own) indicating that, when a particu-
lar word form is activated, formally similar words known to the language-
user are also activated, whatever the language affiliation of the words in
question (at least in respect of languages which are genuinely active in the
individual’s mind). According to this view, ‘the bilingual brain cannot
avoid language conflict, because words from the target and non-target
languages become automatically activated’ (van Heuven, Schriefers,
Dijkstra and Hagoort 2008, p. 2706). A not dissimilar kind of interpretation
can be applied to the language-specific processing cost paid by bilinguals
in the syntactic domain (Hernandez, Bates and Avila 1994). Parallel activa-
tion and language conflict do not of themselves argue for unboundedness;
quite the contrary. While their rhetoric appears sometimes to come close
to claiming cross-linguistic unitariness, Dijkstra and his collaborators in
the end retreat from such a conclusion.
Dijkstra (2003) is obliged to acknowledge, on the basis of his findings,
that individual languages can be at different levels of activation and to
propose a model in which ‘language nodes’ are operative, thus recognising
that the lexical items and processes associated with each of the languages
known to an individual may be activated and/or de-activated as a set. He
and Kroll have the following to say about the context of production:
to speak a word in order to name a pictured object, it is mandatory that
language be specified. Even in the case of highly similar cognate transla-
tions, words in two languages rarely have an identical pronunciation, so
language must be known if performance is to be error-free.
(Kroll and Dijkstra 2005, p. 320)
One notes that whether or not one knows a language is not the only
factor at play. In language reception, assumptions about the language
being used on the basis of general probabilities also come into play with
dramatic effects. Where the expectation is that language x is being spoken
but where, in fact, language y is being used, comprehension may be
impeded, even where both languages are perfectly well known to the
individual concerned, as the example below (from the Finnish psycholo-
gist Elisabet Service) shows:
My sister, while studying in France was once addressed on the street in
Finnish. Only after several attempts by the speaker did she understand her
own native language, the point being that she was expecting French.
I have had a very similar experience trying to make Finnish out of

something that was easy enough to understand when I realised it was

English. (Elisabet Service, personal communication)
The implication of this kind of phenomenon is that the specific nature of

the input alone does not suffice to set the context for tuning the mind to
make sense of even a language which is in principle known, even native.
Somehow or other a ‘switch’ needs to be turned on – in response to a
variety of cognisances – which summons up the whole set of capacities
necessary for coping with a given language.
24.3.2 The relationship of languages in the mind

at the conceptual level
As Athanasoupoulos (Chapter 17, this volume) rightly states, and as men-
tioned above in the introductory section, the most influential accounts of
the mental architecture of multilingualism assume equivalence, fusion,
language-neutrality at the conceptual level (see e.g. de Groot 1992; Kroll
and Stewart 1994; Costa 2005). This idea is, it has to be said, very much in
conflict with the way language-users seem to experience having a variety
of languages available to them. People very often report that they feel that
their ‘take’ on the world changes as they switch from language to
language. Of course, the classic proponent of this perspective was the
early Latin poet Quintus Ennius, who, according to Aulus Gellius, used to
say that because he could speak three languages, Greek, Oscan and Latin,
he had ‘three hearts’ (‘Quintus Ennius tria cordia habere sese dicebat, quod
loqui Graece et Osce et Latine sciret’).
To bring the discussion closer to our own times, we also need to bear in
mind the well-worn truism of modern linguistics (see Lyons 1963, p. 37f.),
according to which every language articulates the world uniquely in terms
of its various structures and consequently in terms of its concepts and
configurations of concepts. (This clearly underlay what Quintus Ennius
was talking about!) What it implies is that, in order to function intelligibly
and comprehendingly in the relevant language communities, users of
multiple languages need to make use of conceptual systems specific to
each of their languages, systems which are of their nature differentiated
from those of their other languages. The reality of a degree of cross-linguistic
conceptual permeability, influence and transfer does not undermine the
other reality of this essential differentiation (see Singleton 2003, 2012).
Athanasoupoulos (Chapter 17) speaks of the need to re-examine the
(in some circles established) notion of a common conceptual base relating
to the languages of multilinguals. He cites in this context the work of
Pavlenko and Jarvis (Pavlenko 2005; Jarvis and Pavlenko 2008). The points
he makes are that most words that are considered ‘translation equivalents’
across languages do not share the same conceptual representation, even
when they denote concrete entities (Ameel et al. 2005), and that there is

increasing evidence from investigations of bilingual cognition of systema-

tic cross-linguistic variation in the conceptual representation of a range of
different domains of experience. His conclusion is that learning a new
language involves creating new concepts and recalibrating existing
concepts.
Let us return to Quintus Ennius. When he said he had three hearts, he
was referring to three centres of intelligence and thought as well as seats of
the emotions. The Latin word cor has a broader set of connotations than
English heart. One notes, however, that the emotional connotation was
nevertheless included, and so it is in connection with current research into
conceptual divergences amongst the languages of the multilingual mind,
which also addresses emotional concepts. Thus, Dewaele (Chapter 22, this
volume) makes mention of Pavlenko’s (2006) analysis of the feedback from
the Bilingualism and Emotions Questionnaire (Dewaele and Pavlenko
2001–2003). Pavlenko found that almost two-thirds of participants
reported feeling like different people when they switched languages.
(See also Grosjean’s interesting words on ‘Emotions in more than one
language’ on his Life as a Bilingual blog: www.psychologytoday.com/blog/
life-bilingual/201108/emotions-in-more-one-language, retrieved 6 August
2014.) The evidence at the emotional level too, then, is for differentiation
between the multilingual’s languages.
24.3.3 The truth about apparent language confusion

A conception of the relationship in the mind of the multilingual’s
language knowledge in terms of an amorphous blur might be considered
to be favoured by evidence that language-users confuse the languages they
have at their disposal. Our earlier discussion of code-switching and trans-
languaging is of relevance here. A superficial view of these phenomena
might see them as favouring a sort of ‘all-in-together’ perspective on
language knowledge. Indeed, evidence of language alternation was once
used as an argument for the complete unitariness of the bilingual’s
language system in early stages. We have noted that code-switching
actually responds to the structural differences between the alternating
languages, which does not look like mish-mash, and that translanguaging
is coordinated and meaningful, not a muddle but an especially
sophisticated feat of steering among the languages and cultures known
to the translanguagers in question.
Especially interesting in any discussion of purported language confu-
sion is the case of cognates which have been borrowed. They are often
re-deployed in the original source language by learners of that language
who have encountered them in the language which borrowed them. For
example, Daulton concludes (2008, pp. 72–76) that Japanese learners of
English expand their resources in English by exploiting their familiarity
with English loanwords in Japanese; but this does not signify a confusion

of languages. It in fact indicates a high degree of language difference

awareness. Japanese learners are no doubt aware of the English-language
origins of many of the English loanwords they use in Japanese, on the basis
of the phonological shapes of the words in question, on the basis of their
written katakana forms (the katakana script being reserved for Western
borrowings), on the basis of having encountered their original forms in
English and/or simply on the basis of general knowledge. When they
exploit their familiarity with such words in their use of English, therefore,
they are not mixing or conflating languages; they are rather making very
sensible use of their knowledge, or perhaps even only their suspicion, that
Japanese borrowed the items in question from the language they are trying
to come to grips with. A similar phenomenon is observable in Polish
learners of English, who make use of any borrowed cognates and shared
borrowed cognates they can lay their hands on (Otwinowska-Kasztelanic
2012).
Finally, we should consider instances where language users are
genuinely unsure which language they are using; surely this counts as
language confusion! A case in point is Singleton’s (1987) study of Philip, an
English-speaking learner of French who also had residual school knowl-
edge of Latin and Irish, plus colloquial knowledge of Spanish gained
during a three-year stay in Spain. When trying to use French, Philip often
made use of Frenchified Spanish expressions. For instance, when trying to
express ‘although’ he produced a Frenchified version of the Spanish
expression for ‘although’ – aunque [a’ɔ̃ke]. From his introspections on his
use of this form, it emerged that he did not know whether he had
borrowed it from Spanish or Latin. On the other hand, in general, Philip
appeared to be fairly focused and self-aware about his borrowing from
Spanish. He mentioned that he knew Spanish and French to be historically
related and he reported deliberately making use of Spanish expressions to
fill gaps in his French. Indeed, it turned out that Spanish was the privileged
source of transferred expressions in his French, which symptomises a
well-motivated transfer strategy. As for not knowing whether aunque was
Latin or Spanish, given his rather limited command of Spanish and his
only vestigial knowledge of Latin, we should probably not make too much
of this – especially since Latin and Spanish in fact have many expressions
in common.
De Angelis (2007, pp. 65–67) discusses such uncertainties in a
comparable fashion. For example, she cites a French-Canadian learner of
Italian with prior knowledge of Spanish who when speaking Italian used
the Spanish word for ‘table’ – mesa – instead of the Italian word tavolo,
believing mesa to be Italian. Also cited by de Angelis is an English learner of
Spanish with previous knowledge of Italian who was familiar with two
words for ‘money’ – dinero (Spanish) and soldi (Italian) – but was unable to
say which word belonged to which language. A third example mentioned
by de Angelis occurs in a study (Bardel and Lindqvist 2007) in which one of

the participants, on uttering ahora, admitted to being unsure whether it

was Italian or Spanish. De Angelis takes a similar line in this connection to
that proposed regarding Philip’s case. She suggests that ‘language tags or
cues may have different strength values at different stages of the acquisi-
tion process, and strengths may vary depending on changes in language
proficiency over time’ (de Angelis 2007, p. 86). This is a very different kind
of account from that which completely dismisses the idea of boundaries in
the mind between the languages known to an individual. Noteworthy also
is the fact that in the above examples, the additional languages in question
are Spanish and Italian, which, like Latin and Spanish, to an extent share
lexicons. Such confusions would certainly not have arisen if the languages
in question had been Hungarian and Chinese.
One has to conclude that there are indeed cases, at particularly
impoverished levels of proficiency, of multilingual individuals being
confused about which language some of their utterances belong to –
especially where their languages are closely related. One has also to
say, again, however that the exigencies of multilingual communica-
tion mean that confusion cannot come anywhere close to ‘ruling the
roost’. For multilingual communication to work (and it does work on
the whole) the participants must be acutely sensitive to what will and
will not be understood by their interlocutors, and must navigate
amidst the various languages at their disposal adroitly and precisely.
24.4 Envoi
For my envoi, I should like to refer to the work of Janet Werker on early
bilingualism (2012). Werker points out that the infant facing into the
process of language development has to deploy his/her
perceptual knowledge of the rhythmical properties of the . . . language, of
the speech sound categories that distinguish one possible word from
another, and of the sequences of sounds that are allowable within a
word and/or the statistical learning of other cues to segmentation.
(Werker 2012, p. 50)
Only on this basis, she says, can the child isolate different words and
structures and map them on to meaning. The child who grows up in an
environment involving more than one language, she goes on to point out,
has to master the rhythmical properties, the phonetic categories, the
phonotactic regularities, the word order patterns, the lexis–concept
configuration and the conceptualisation of the world of each language.
What is more, the infant bilingual must do this, she states, without
interlingual confusion.
My comment, of course, is that, in this connection, what is true for the
infant bilingual is true for all users of more than one language.

References
Ameel, E., Storms, G., Malt, B. and Sloman, S. 2005. ‘How bilinguals solve
the naming problem’, Journal of Memory and Language 53 (1), 60–80.
Benjamins.
Bardel, C. and Lindqvist, C. 2007. ‘The role of proficiency and psycho-
typology in lexical cross-linguistic influence: a study of a multilingual
learner of Italian L3’, in M. Chini, P. Desideri, M. E. Favilla and G.
Pallotti (eds), Atti del VI Congresso Internazionale dell’ Associazione Italiana
di Linguistica Applicata, Napoli, 9–10 febbraio 2006, pp. 123–145. Perugia:
Guerra Editore.
Bybee, J. L. 1988. ‘Morphology as lexical organization’, in M. Hammond
and M. Noonan (eds), Theoretical Morphology, pp. 119–141. San Diego:
Academic Press.
Clark, E. 2009. First Language Acquisition, second edition. Cambridge
University Press.
557–591.
Cook, V. J. 1999. ‘Going beyond the native speaker in language teaching’,
TESOL Quarterly 33 (2), 185–209.
Cook, V. J. 2003. ‘Introduction: the changing L1 in the L2 user’s mind’, in
V. J. Cook (ed.), Effects of the Second Language on the First, pp. 1–18. Clevedon:
Cook, V. J. and Singleton, D. 2014. Key Topics in Second Language Acquisition.
Costa, A. 2005. ‘Lexical access in bilingual production’, in J. F. Kroll and
A. M. B. de Groot (eds), pp. 308–325.
Daulton, F. E. 2008. Japan’s Built-in Lexicon of English-based Loanwords.
De Bot, K. 2004. ‘The multilingual lexicon: modelling selection and
control’, International Journal of Multilingualism 1 (1), 17–32.
Cognition 10 (1), 7–21.
De Bot, K. and Schreuder, R. 1993. ‘Word production and the bilingual
lexicon’, in R. Schreuder and B. Weltens (eds), The Bilingual Lexicon,
Phonology, Morphology and Meaning, pp. 389–412. Amsterdam: North-
Holland.

De Houwer, A. 1995. ‘Bilingual language acquisition’, in P. Fletcher and

B. MacWhinney (eds), The Handbook of Child Language, pp. 219–250.
Malden, MA: Blackwell.
Deuchar, M. and Quay, S. 2000. Bilingual Acquisition: Theoretical Implications of
a Case Study. Oxford University Press.
Dijkstra, T. 2003. ‘Lexical processing in bilinguals and multilinguals: the
word selection problem’, in J. Cenoz, B. Hufeisen and U. Jessner (eds), The
Multilingual Lexicon, pp. 11–26. Dordrecht: Kluwer Academic.
Fabbro, F. 1999. The Neurolinguistics of Bilingualism: An Introduction. Hove:
Psychology Press.
Fabbro, F. 2001. ‘The bilingual brain, part 2: cerebral representation of
languages’, Brain and Language 79 (2), 211–222.
Fabbro, F. 2002. ‘The neurolinguistics of L2 users’, in V. J. Cook (ed.),
pp. 167–195.
Garcia, O. and Li Wei. 2014. Translanguaging: Language, Bilingualism and
Education. Basingstoke: Palgrave.
Genesee, F., Nicoladis, E. and Paradis, J. 1995. ‘Language differentiation in
early bilingual development’, Journal of Child Language Development 22 (3),
611–631.
Gentner, D. 2010. ‘Bootstrapping the mind: analogical processes and
symbol systems’, Cognitive Science 34 (5), 752–775.
Gil, M. and Goral, M. 2004. ‘Non-parallel recovery in bilingual aphasia:
effects of language choice, language proficiency and treatment’,
International Journal of Multilingualism 8 (2), 191–219.
Green, D. W. 1986. ‘Control, activation and resource: a framework and a
model for the control of speech in bilinguals’, Brain and Language
27 (2), 210–223.
Grosjean, F. 1982. Life with Two Languages: An Introduction to Bilingualism.
Cambridge, MA: Harvard University Press.
Grosjean, F. 1992. ‘Another view of bilingualism’, in R. Harris (ed.),
Cognitive Processing in Bilinguals, pp. 51–62. Amsterdam: North-
Holland.
Harris, R. 1998. An Introduction to Integrational Linguistics. Kidlington:
Elsevier.
Hernandez, A. E., Bates, E. and Avila, L. X. 1994. ‘On-line sentence inter-
pretation in Spanish–English bilinguals: what does it mean to be “in
between”?’ Applied Psycholinguistics 15 (4), 417–446.
Huang, Y. T. and Pinker, S. 2010. ‘Lexical semantics and irregular inflec-
tion’, Language and Cognitive Processes 25 (10), 1411–1461.
Kanno, Y. 2003. Negotiating Bilingual and Bicultural Identities: Japanese Returnees
betwixt Two Worlds. Mahwah, NJ: Lawrence Erlbaum.

Kecskes, I. and T. Papp 2000. Foreign Language and Mother Tongue. Mahwah,
Kouritzin, S. G. 1999. Face(t)s of First Language Loss. Mahwah, NJ: Lawrence
Erlbaum.
Psycholinguistic Approaches. New York: Oxford University Press.
Kroll, J. F. and Dijkstra, T. 2005. ‘The bilingual lexicon’, in R. B. Kaplan (ed.),
The Oxford Handbook of Applied Linguistics, pp. 301–321. Oxford University
Press.
picture naming: evidence for asymmetric connections between bilin-
gual memory representations’, Journal of Memory and Language 33 (2),
149–174.
Lanza, E. 1997. Language Mixing in Infant Bilingualism: A Sociolinguistic
Perspective. Oxford: Clarendon Press.
Li Wei 2011. ‘Moment analysis and translanguaging space: discursive
construction of identities by multilingual Chinese youth in Britain’,
Lyons, J. 1963. Structural Semantics: An Analysis of Part of the Vocabulary of Plato,
vol. 20 of Publications of the Philological Society. Oxford: Blackwell.
Macrory, G. 2006. ‘Bilingual language development: what do early years
practitioners need to know?’, Early Years 26 (2), 159–169.
Meisel, J. M. 2008. ‘Child second language acquisition or successive first
language acquisition?’, in B. Haznedar and E. Gavruseva (eds), Current
Trends in Child Second Language Acquisition, pp. 55–80. Amsterdam: John
Benjamins.
Myers-Scotton, C. 2003. ‘Code-switching: evidence of both flexibility and
rigidity in language’, in J.-M. Dewaele, A. Housen and Li Wei (eds),
Bilingualism: Beyond Basic Principles, pp. 189–203. Clevedon: Multilingual
Matters.
Nicoladis, E. 1998. ‘First clues to the existence of two input languages:
pragmatic and lexical differentiation in a bilingual child’, Bilingualism:
Language and Cognition 1 (2), 105–116
Nicoladis, E. and Secco, G. 2000. ‘The role of a child’s productive vocabulary
in the language choice of a bilingual family’, First Language 58 (1), 3–28.
Otwinowska-Kasztelanic, A. 2012. ‘Plurilingualism and Polish learners of
English’, Linguarum Arena 3, 37–52.
Paradis, M. 1981. ‘Neurolinguistic organization of a bilingual’s two lan-
guages’, in J. Copeland and P. Davis (eds), The Seventh LACUS Forum,
pp. 486–494. Columbia, SC: Hornbeam Press.
Paradis, M. and Goldblum, M. C. 1989. ‘Selective crossed aphasia in a
trilingual aphasic patient followed by reciprocal antagonism’, Brain
Groot (eds), pp. 433–453.

Pavlenko, A. 2006. ‘Bilingual selves’, in A. Pavlenko (ed.), Bilingual Minds:

Emotional Experience, Expression and Representation, pp. 1–33. Clevedon:
Pettito, L. A., Katerelos, M., Bronna, G., Levy, K. G., Tétreault, K. and
Ferraro, V. 2001. ‘Bilingual signed and spoken language acquisition
from birth: implications for the mechanisms underlying early bilin-
gual language acquisition’, Journal of Child Language Acquisition 28 (2),
453–496.
Pfenninger, S. and Singleton, D., forthcoming. Beyond Age Effects: Facets, Facts
and Factors of Foreign Language Instruction in a Multilingual State. Bristol:
Quay, S. 1995. ‘The bilingual lexicon: implications for studies of language
choice’, Journal of Child Language 22 (2), 369–387.
Safont Jordà, M. P. 2005. Third Language Learners: Pragmatic Production and
Awareness. Clevedon: Multilingual Matters.
Schelleter, C., Sinka, I. and Garman, M. 1997. ‘Latvian/English and German/
English bilingual acquisition: new light on Universal Grammar’, in
Proceedings of the First International Symposium on Bilingualism, pp. 442–452.
University of Vigo.
Schwyter, J. R. 2011. ‘ “Me talk funny”: a stroke patient’s personal account’,
English Today 27 (4), 49–52.
Singleton, D. 1987. ‘Mother and other tongue influences on learner
French’, Studies in Second Language Acquisition 9 (3), 327–345.
Singleton, D. 2003. ‘Perspectives on the multilingual lexicon: a critical
synthesis’, in J. Cenoz, B. Hufeisen and U. Jessner (eds), The Multilingual
Lexicon, pp. 167–176. Amsterdam: Kluwer Academic.
Singleton, D. 2012. ‘Multilingual lexical operations: keeping it all
together . . . and apart’, in J. Cabrelli Amaro, S. Flynn and J. Rothman
(eds), Third Language Acquisition in Adulthood, pp. 95–114. Amsterdam: John
Benjamins.
Singleton, D., Fishman, J., Aronin, L. and Ó Laoire, M. 2013. Current
Multilingualism: A New Linguistic Dispensation. Berlin: Mouton De Gruyter.
Sinka, I., Garman, M. and Schelleter, C. 2000. ‘Early verbs in bilingual
acquisition’, Reading Working Papers in Linguistics 4, 175–187.
Sinka, I and Schelleter, C. 1998. ‘Morphosyntactic development in
bilingual children’, Journal of Bilingualism 2 (3), 301–326.
Stemberger, J. and MacWhinney, B. 1988. ‘Are inflected forms stored in the
lexicon?’, in M. Hammond and M. Noonan (eds), Theoretical Morphology,
pp. 101–116. New York: Academic Press.
Toolan, M. 2008. ‘Introduction: language teaching and integrational
linguistics’, in M. Toolan (ed.), Language Teaching: Integrational Linguistic
Approaches, pp. 1–23. London: Routledge.
Van Heuven, W. J., Schriefers, H., Dijkstra, T. and Hagoort, P. 2008.
‘Language conflict in the bilingual brain’, Cerebral Cortex 18 (11),
2706–2716.

Volterra, V. and Taeschner, T. 1978. ‘The acquisition and development of

language by bilingual children’, Journal of Child Language 5 (2), 311–326.
Werker, J. 2012. ‘Perceptual foundations of bilingual acquisition in
infancy’, Annals of the New York Academy of Sciences 1251, 50–61.
Whitaker, H. 1978. ‘Bilingualism: a neurolinguistics perspective’, in
W. Ritchie (ed.), Second Language Acquisition Research: Issues and
Implications, pp. 21–32. New York: Academic Press.
Zhang, H.-Q. 2006. Bilingual Sibling Interaction: A Case Study of Linguistic
Features and their Social Functions. MPhil dissertation. Trinity College
Dublin.

25
Questions of
multi-competence: a
written interview
Guillaume Thierry
These questions were posed by the editors and contributors after

Guillaume Thierry had read drafts of the chapters.
25.1 General questions
1 Does bilingualism research reveal anything about the human mind that we
wouldn’t otherwise know if we were studying only monolinguals?
I believe that time for speculative, contemplative research is running out.

Not because scientists aren’t interested in those topics and questions that
do not have direct ‘real-world’ applications but simply because funding is
progressively more and more geared towards return on investment. Thus,
the fact that research on bilingualism is now flourishing is itself a demon-
stration of its value and the fact it has a lot to say about the human mind
that we would not know if we only studied monolinguals.
First, the prevalence of bi- and multilingualism in the modern world is
no longer a fact known only to experts. Now faced with a global economy,
increasingly diverse and integrative legal and political negotiations,
and cultural fusion underpinned by ethnic mixing – to take only a few
examples – it is natural that multi-competence, as defined in this book,
becomes a generalised and unavoidable phenomenon. If the phenomenon
is so obviously prevalent and only set to become more widespread, it
cannot be avoided as a topic of research in its own right.
I thank Vivian Cook for his invitation to prompt this contribution by generating a list of questions and for his comments on
a previous version of the text.

522 GUILLAUME THIERRY
Second, the answer to the question seems obvious if one considers

that bi- and multilingualism combine language ability with the highly
sophisticated executive mechanisms required to operate two or more
language inputs and outputs simultaneously. Linguistics, psycholinguis-
tics and cognitive neuroscience have taught us a great deal about the
mechanisms underpinning language comprehension, representation,
and production on the one hand and about attention, memory, emotion,
executive control on the other, but the study of bilingualism and multi-
lingualism offers a unique combination of all these in a single mind. This
is not to say that language does not interact fully with all other domains
of cognition in monolinguals as well (this is an issue I will get back to
later because I disagree with the proposal that the multi-competence
approach is novel in arguing that language interacts with other domains
of human cognition). Thus, the cognitive mechanisms underpinning bi-
and multilingualism and the corresponding behaviours manifested sim-
ply cannot be generalised from the monolingual operating paradigm. As
famously pointed out by Grosjean (1989), a bilingual is not simply two
monolinguals coexisting in the same mind. It follows that specific cogni-
tive operations and systems are in place to allow an individual to speak
one language without suffering apparent interference from the other(s)
they speak.
Third, I have come to take the view that bilingualism is fertile ground
for the exploration of human consciousness and, de facto, simultaneously,
the unconscious, in ways that ‘monolingual research’ – if there is such a
thing – is profoundly underpowered. If language is amongst the most
sophisticated products of the human brain, managing two or more
systems of sound forms, orthographic representations, phono- and
orthotactic rules, terminologies, syntactic frames, expressions, prag-
matic schemas and metaphorical references (to name only a few of the
sub-processes) calls upon other cognitive domains in ways which have
to be qualitatively and not merely quantitatively different. If you agree
that the complexity and level of cognitive interaction required by bi- and
multilingual operation is unmatched, it becomes obvious that it deserves
its own research stream.
2 Psychologists publishing on language topics in psychological journals rarely

refer to applied linguistic/multilingualism research. Is that due to ignorance?
contempt? incompatibility of research methods or questions?
Frankly, all three. My feeling (and let’s be clear, this is only my impression
and I do not have hard evidence or statistics to offer here) is that:
(i) Psychologists rarely delve into the meanders of applied linguistics

and certainly consider that they have enough to read and integrate
from their own field not to ‘waste’ time reading scholarly linguistic
works. To be fair, I sense some kind of asymmetry because I have met

Questions of multi-competence: a written interview 523
and collaborated with several linguists who have engaged in a

masterful transition to experimental psychology (e.g., Panos
Athanasopoulos, now at the University of Lancaster) or who have
recognised the power to be gained from scientific evidence when
dealing with ‘humanistic’ questions (e.g., Phil Davis at the University
of Liverpool).
(ii) Methods are sometimes non-existent, often poor, and almost always
statistically untestable in applied linguistics.
(iii) There is no real incompatibility regarding questions and in fact
there should certainly be none. Sharing questions and attempting
to find a terrain of mutual understanding for collaboration is the
essence of inter-disciplinary research. In an ideal world, bi- and
multilingualism research should involve the applied linguist as the
question setter, the psychologist as the experiment designer and
the cognitive neuroscientist as data collection and analysis expert.
Better still, the three should reach a level of communication and
mutual understanding sufficient to conceive and implement all
three facets of the process together, whilst being able to appreciate
each other’s perspective. In my experience, such collaborative efforts
are extremely productive and can lead to groundbreaking new
insights (e.g. Thierry et al. 2008; Thierry et al. 2009; Athanasopoulos
et al. 2010; Keidel et al. 2013).
3 How do you see the relationship between language and the rest of the mind?
Is language something of its own or nothing more than the intersection of
diverse ‘non-language’ processes? How do you locate bilingualism in these
terms?
It will already be clear from my previous replies that the relationship

between language and the rest of the mind should be considered as
profoundly intimate and bi-directional. Together with scholars like
Friedemann Pulvermüller (e.g. 1999) and proponents of connectionism
and the Hebbian theory (1949), I see the human brain as a vastly inter-
connected network in which functional sub-networks are characterised
by cells firing in coincidence, rather than as a juxtaposition of encapsu-
lated modules that are anatomically and/or functionally distinct (see e.g.
Démonet et al. 2005). In fact, I would go as far as saying that making a
distinction between language and the rest of the mind is meaningless.
Making such a distinction implies that language and mind are two ensem-
bles that can be delimited, as if one could draw a line between the two, or
indeed trace a line around language within the mind. This is misleading
both from an anatomical and functional viewpoint. First, there is no such
thing as a language-specific brain region (see e.g. Price et al. 2005 for a
discussion of this issue in the domain of speech). It has been widely
shown that the areas of the cerebral cortex, inner brain ganglia and
the cerebellum involved in language processing are also activated by

numerous non-verbal auditory and visual processes. Even though some

recent studies have offered evidence in favour of what may be construed
as micro-selectivity in classic brain regions such as Broca’s area (e.g.
Fedorenko et al. 2012), the spatial resolution of functional neuroimaging
is so far largely insufficient to establish neuroanatomical selectivity and
certainly no convincing evidence has yet been obtained at the scale of
brain regions. Second, there is no such thing as a cognitive operation
impermeable to or wholly independent of language processing and vice
versa. Work by our group at Bangor has shown that distinctions made by
terminology in a given language constrain perception (Thierry et al. 2009;
Athanasopoulos et al. 2010) and early visual categorisation (Boutonnet
et al. 2013; Flecken et al. 2015). In collaboration with Yanjing Wu, we
found that a bilingual context of operation prompts a different modus
operandi in terms of executive functioning, such that mere exposure to
words in the two languages of fluent bilinguals increases their ability to
control interference in a purely non-verbal task (the arrow flanker task,
Wu and Thierry 2013). More recently, we have even found evidence for
profound interactions between language of operation (manipulated
merely by presenting feedback words in either of two languages) and
risk-taking when Chinese–English bilinguals engage in gambling (Gao
et al. 2015).
As regards defining language as ‘no more than the intersection of
diverse “non-language” processes’, I am not sure what this means.
Perhaps this is an allusion to the idea that language may be an emer-
gent phenomenon, essentially arising, along with conscious experi-
ence, from layers of mental processing that are wholly unconscious.
If this is the idea put forward here, I must say that I am very tempted
to conceive language and consciousness as emergent phenomena
essentially incapable of any ‘top-down’ impact on volition and deci-
sion-making.
25.2 Premises of multi-competence (Chapter 1)
4 Comment on Premise 1: Multi-competence concerns the total system for all

languages (L1, L2, Ln) in a single mind or community and their inter-
relationships.
This definition may be too inclusive to be practical. It carries the risk of

making the phenomenon essentially impossible to apprehend, study,
and thus, understand. This to me resembles the definition of a whole
discipline rather than a particular phenomenon. In fact I would be
reasonably happy with this premise as a definition for sociolinguistics,
even though I am aware that sociolinguists probably would strongly
object. It may be useful to explain how multi-competence distinguishes

itself from sociolinguistics. In other words, it may be useful to better

define what it is not rather than what it is.
5 Comment on Premise 2: Multi-competence does not depend on the mono-

lingual native speaker.
Perhaps this premise is partly in contradiction with Premise 1. If a com-

munity has a strong contingent of monolingual speakers (as is the case in
Wales for instance), is it fair to say that the stance of the monolingual
speaker and purely monolingual language use have no bearing on the
linguistic make-up of the community?
6 Comment on Premise 3: Multi-competence affects the whole mind, i.e. all

language and cognitive systems, rather than language alone.
From my reply to questions 1 and 4 above, Premise 3 is a given, since

nothing can affect language alone in my view. If researchers in
linguistics and sociolinguistics need a wake-up call regarding the
inter-relations of language with other types of cognitive operations
implemented in the human brain, then the premise is warranted as
a reminder.
25.3 Other questions
7 What would you see as the role of the monolingual native speaker in
bilingualism research?
This is a good question. Imagine asking a physiologist working on the

digestive system of the bat what she could use birds or rodents for. Of
course a comparison between the two (three?) is tempting. But is it
really meaningful? It could be argued that such comparison is ludi-
crous, that, in fact, comparing bilinguals with monolinguals is like
comparing apples and oranges. It seems that studies on bilingualism
so far have established results that are for the most part idiosyncratic
to bilingualism and shed little or no light on monolingual processing.
Research on monolinguals remains very interesting in its own right,
since it is a case of ‘pure’ interactivity between language processes and
other cognitive mechanisms. In sum, the problem seems to come from
the assumption that bilinguals should be systematically compared to
monolinguals, who become de facto the default ‘control’ group. My
view, however, is that we should walk away from group comparisons
and wherever possible use within-subject experimental designs in
which bilinguals are their own controls (just as in monolingual-based
experiments on language). In my group, we have successfully engaged
in such experiments recently, even though such designs cannot
(directly) answer questions of the kind: ‘How do bilinguals differ from

monolinguals?’ (Wu and Thierry 2013; Vaughan-Evans et al. 2014; Gao

et al. 2015; Ellis et al. 2015).
25.4 Research methods
8 What can a psychological/neuroscience approach reveal about bilingualism

that wouldn’t otherwise be revealed by say methods coming solely from
linguistics, or observations of language use?
Unfortunately, a thorough reply to this question would probably require a

rather extensive essay. An analogy may help here. Consider the following
question, with the same implications: What can the crime scene/lab analy-
ses reveal about a murder case that wouldn’t otherwise be revealed by just
talking to the relations of the victim and the suspects? Whilst the perspec-
tive, tools and methods of psychology and neuroscience are entirely dif-
ferent from those of linguistics, these approaches are rather obviously
complementary in nature. What the analogy hopefully makes clear is
that the different approaches will fail to explain the phenomenon under
study if one or the other is ignored. By this I mean that applied linguistics
and sociolinguistics need psychology and neuroscience as much as the
reverse.
9 Multi-competence research has often used ‘simple’ experiment tasks such

as five-point-scale judgments and word associations? How much can these
actually reveal about cognitive processes among speakers of more than one
language? What would you prefer?
This question highlights the existence of various methodological

approaches and measurement tools to study a particular phenomenon
and hints at the fact that important questions can be addressed without
resorting to highly technical methods. There is a considerable array of
measurement tools available to researchers in the fields of bi- and multi-
lingualism. On one end of the spectrum, I would place introspection,
which I understand is used in fields of the humanities such as philosophy
and cognitive linguistics, for instance. Towards the other extreme I would
place outputs from computational modelling or single cell recordings
from the human cortex. The latter are purely quantitative, elementary
measurements collected in response to the manipulation of a given (or
several) experimental factor(s) with very limited degrees of freedom.
Between such two extreme scenarios, essentially everything is possible:
one can choose an interview format (‘How do you manage your two
languages on a daily basis? Is it harder for you to switch from your native
to your second language than the reverse? Have you noticed intrusions
from your native language when you speak you second language?’ etc.) as
is the case in Cognitive Discourse Analysis (e.g. Tenbrink et al. 2012),
questionnaires inviting yes or no answers, confidence ratings, or Likert

scales. One can choose to use a programmed experiment registering accu-

racy in a number of trials, perhaps asking for speeded responses, recording
reaction times (using a voice key or a response box). One can also elect
to take physiological measurements such as galvanic skin response, eye-
tracking measurements, electrophysiological data, cerebral blood flow
variation estimates, etc. Naturally all these measurements are not equiva-
lent in terms of reliability, and thus they do not lend themselves to
statistical evaluation in the same way. In my view the key question is not
how much these methods can reveal about multilingual cognition (is it
possible to know?), but rather what sort of questions they can and cannot
answer. A method of investigation should be chosen to address a particular
question.
In my research I use mostly event-related potentials (ERPs), because
this method can provide reliable insights about unconscious information
processing and test access to specific mental representations in the
absence of awareness on the part of the participant. In 2013, with my
colleague Yanjing Wu, we set out to determine if Chinese–English bilin-
guals would unconsciously and automatically access translation equiva-
lents of English words in their native language Chinese (as we established
before, e.g. Thierry and Wu 2007; Wu and Thierry 2010; Wu and Thierry
2012) even when there was no requirement and in fact no useful informa-
tion to be gained from reading words in the experiment. Since we wanted
to know whether participants would translate uninformative words, we
had to make reading optional and thus could not use ERPs, because this
method relies on the presentation of stimuli at fixation, and since reading
is automatic, words would have been read in any case. We thus chose the
methods – eye-tracking – to suit the question we were asking, because it
was the best method to answer that particular question. And indeed, we
found that Chinese–English bilinguals accessed the translations of English
words in their native language even when they have no requirement to
read them – or even look at them (Wu et al. 2013). In the same vein,
Athanasopoulos and colleagues recently established – using purely beha-
vioural measures of choice – that the two languages of a bilingual are
coactive and compete for cognitive resources in bilinguals, even when
the main task at hand is visual and does not trivially rely on language
processing (Athanasopoulos et al. 2015).
In sum, to investigate the potential advantages of maintaining one’s
bilingual status in a particular geographical location or a particular sector
of professional activity, it may be adequate to interview a sample of
screened workers and ask them simple questions regarding well-being at
work, opportunities offered to them, salary, etc. However, if the goal is to
understand whether bilinguals are better able to dismiss irrelevant or
conflicting information in conditions where there is competition for
attentional resources, it would be more suitable to use an experimental
task involving non-verbal stimuli and take measurements unlikely to be

knowingly or strategically biased by the participant. Now imagine asking

a bilingual whether she suffers interference from her native language
when working in her second language. And, in the same vein, imagine
submitting the workers in the former case to non-verbal speeded reaction
time tasks. Not only would the method seem unsuited to the questions
asked, but predictions would also be essentially impossible, and
hypothesis-testing would break down, opening the way to speculation
and misinterpretation.
10 A huge amount of psychological research is based on samples of American

college students whose language profile, i.e. other languages known, is never
mentioned nor taken into account. Is that a problem?
This is commonly known in scientific research as the problem of

uncontrolled variables, which are often associated with an experimen-
tal bias. If a relevant characteristic of the group studied is not mea-
sured when it is likely to have an impact on the results, then the
study’s conclusions are invalid or of low reliability. In this case, salva-
tion can only come from replication, the importance of which is
largely underestimated nowadays. If a substantial number (and a for-
tiori the majority) of independent studies tackling a particular question
converge in their findings and conclusions, the conclusions are likely
warranted through reproducibility. There are statistical tools to evalu-
ate the degree of reproducibility of a result and, more importantly,
there are tools to evaluate possible publication biases (see e.g. de Bruin
et al. 2015).
25.5 Reactions to statements made by different

contributors
How do you react to these statements in different chapters?
11 ‘We would expect that all that we currently take to be a “given” in psycholin-
guistic research would have to be recast as applicable only to “monolingual”
language processing.’ (Vaid and Meuter, Chapter 4, p. 92)
I disagree with the view that all psycholinguistic research to this day is
exclusively based on the study of the monolingual case. There is a strong
(and growing) base of psycholinguistic and neurolinguistic literature
dedicated to bilingualism, which derives from the observation and testing
of bilinguals – not monolinguals. Some pretty well-established models
of language processing (e.g. the Revised Hierarchical Model, Kroll and
Stewart 1994; the Bilingual Interactive Activation model, Dijkstra and
van Heuven 2002) deal with bilingual processing and have been built on
adequate empirical data.

12 ‘What this boils down to is that there are no separate languages but only
discourses.’ (de Bot, Chapter 6, p. 131)
Whilst I understand the desire of the author of this statement to capture

the extraordinary richness and diversity of language make-ups, that is,
how individuals vary from one another in the way their languages interact,
combine and produce an essentially unique mosaic of expressive profiles, I
find the statement too radical or, at the very least, ambiguous. The state-
ment, offered as an incisive summary of the emergentist view, captures
well some aspects of language development in bilinguals, but, considered
out of context and applied to languages as abstract systems, i.e. as theore-
tical entities existing however transiently, independently of the cognitive
system that handles them, it is a frontal provocation of literary scholars
and language conservation experts. The ‘langue française’, for instance,
has manifestly evolved since the eighteenth century, but still it is abso-
lutely clear how it differs in many ways – and thus is separate – from the
English language. The fact that these two languages coexist in my mind,
and differently so in all bilinguals who can use them, does not question the
existence of the two systems of rules and representations that underpin
them. Individuals who can be categorised as monolinguals can tell which
words and structures of any given utterance belong to the language(s) they
know, unless they have not yet achieved a certain level of proficiency. In
fact the very concept of proficiency would be meaningless if languages
were not separate. Languages obviously follow sets of rules at all levels of
representations, and are different from one another. However, I under-
stand that the question of separation discussed here fundamentally con-
cerns the cognitive system underpinning language competence and, once
ambiguity on this front is set aside, I agree that languages are not separate
in the brain, or the mind. Indeed, much data collected in comprehension
and production support the view of language non-selective access to
representations in both comprehension and production (e.g. Thierry and
Wu 2007; Wu and Thierry 2010; Wu and Thierry 2012; Spalek et al. 2014;
Vaughan et al. 2014).
13 ‘[T]he even greater challenge of using concepts like language and

grammar, whose usage is so established and thus their meanings so
entrenched that changing them to mean something else may be impos-
sible, despite our repeated contestations that they be changed.’ (Hall,
Chapter 9, p. 195)
Such a change should be possible and even desirable if a good alternative

were offered (that is, one that is functional, useful, easy to relate to, and
more importantly, underpinned by a reality). How should the concept of
language be changed?

14 ‘[O]ne conclusion can safely be drawn on the basis of the above discussion:
language cannot easily be banned from non-verbal tasks.’ (de Groot,
Chapter 12, p. 270)
In my view language cannot be banned, full stop. Reciprocally, can the

human mind operate at any time or any context without the involvement
of memory, attention or executive control? Thus, non-verbal operations
cannot be banned from language, in the same way as language cannot be
banned from non-verbal operating contexts. Language and non-verbal
cognition are exquisitely and intrinsically intertwined. This is not to say
that non-verbal representations do not exist. Intuitively, I don’t need to
know the name of a fruit to eat it if I am hungry and inclined to try. I can
see and appreciate the indescribable orange-pink colour of a sunset when I
have no adjective in my language(s) to qualify it (even though I concede
that ‘orange-pink’, or ‘indescribable colour’, may do the trick). I have only
a word or two to describe my feelings for my children, my brother, my
parents, my muse, and I certainly feel for them in different ways, but
language fails miserably when it comes to qualifying or conveying such
non-verbal representations. This is not to say that there aren’t functional
network in the brain more attuned to verbal than non-verbal processing
(Démonet et al. 2005) and reciprocally (Thierry et al. 2003; Thierry et al.
2006). In sum, I would argue that verbal processing and non-verbal proces-
sing are as separable in the human brain as the two languages of a
bilingual.
15 ‘For multilingual communication to work (and it does work on the whole)

the participants must be acutely sensitive to what will and will not be under-
stood by their interlocutors, and must navigate amidst the various languages
at their disposal adroitly and precisely.’ (Singleton, Chapter 24, p. 515)
One could hardly disagree with this statement. This being said, I
wonder if there are any differences between the multilingual and
monolingual contexts as regards this issue? The proposal given here
applies generically to human communication and thus applies equally
to the case of monolingual exchanges. Why and how should this
statement specifically apply to multilingualism? I note also that the
statement assumes that languages – as abstract entities – are separate
since multilingual communication requires navigation amidst them.
References
Athanasopoulos, P., Bylund, E., Montero-Melis, G., Damjanovic, L.,

Schartner, A., Kibbe, A., Riches, N. and Thierry, G. 2015. ‘Two languages,
two minds: flexible cognitive processing driven by language of opera-
tion’, Psychological Science 26, 518–526.


Boutonnet, B., Dering, B., Viñas-Guasch, N. and Thierry, G. 2013. ‘Seeing
objects through the language glass’, Journal of Cognitive Neuroscience 25,
1702–1710.
De Bruin, A., Treccani, B. and Della Sala, S. 2015. ‘Cognitive advantage in
bilingualism: an example of publication bias?’, Psychological Science 26,
99–107.
Démonet, J. F., Thierry, G. and Cardebat, D. 2005. ‘The renewal of the
neurophysiology of language: functional neuroimaging’, Physiological
Reviews 85, 49–95.
Dijkstra, T. and van Heuven, W. J. B. 2002. ‘The architecture of the bilingual
word recognition system: from identification to decision’, Bilingualism:
Ellis, C., Kuipers, J., Thierry, G., Lovett, V., Turnbull, O. H. and Jones, M. W.
2015. ‘Language and culture modulate online semantic processing’,
Social, Cognitive and Affective Neuroscience. doi: 10.1093/scan/nsvo28.
Fedorenko, E., Duncan, J. and Kanwisher, N. 2012. ‘Language-selective and
domain-general regions lie side by side within Broca’s area’, Current
Biology 22, 2059–2062.
Flecken, M., Athanasopoulos, P., Kuipers, J. R. and Thierry, G. 2015. ‘On the
road to somewhere: brain potentials reflect language effects on motion
event perception’, Cognition 141, 41–51.
Gao, S., Zika, O., Rogers, R. and Thierry, G. 2015. ‘Second language feed-
back abolishes the “hot hand” effect during even-probability gambling’,
The Journal of Neuroscience 35, 5983–5989.
Grosjean, F. 1989. ‘Neurolinguists, beware! The bilingual is not two
Hebb, D. 1949. The Organization of Behavior. New York: Wiley.
Keidel, J. L., Davis, P. M., Gonzalez-Diaz, V., Martin, C. D. and Thierry, G.
2013. ‘How Shakespeare tempests the brain: neuroimaging insights’,
Cortex 49, 913–919.
Kroll, J. F. and Stewart, E. 1994. ‘Category interference in translation and pic-
ture naming: evidence for asymmetric connections between bilingual
memory representations’, Journal of Memory and Language 33, 149–174.
Price, C., Thierry, G. and Griffiths, T. 2005. ‘Auditory processing specific to
speech: where is it?’ Trends in Cognitive Science 9, 271–276.
Pulvermüller, F. 1999. ‘Words in the brain’s language (Target Article’,
Behavioral and Brain Sciences 22, 253–279.
Spalek, K., Noriko Hoshino, N., Wu, J. Y., Damian, M. and Thierry, G. 2014.
‘Speaking two languages at once: unconscious native word form access
in second language production’, Cognition 133, 226–223.
Tenbrink, T., D’Odorico, T., Hertzberg, C., Mazman, G., Meneghetti, C.,
Reshöft, N. and Yang, J. 2012. ‘Tutorial report: understanding spatial

thought through language use’, Journal of Spatial Information Sciences 5,

107–114.
Thierry, G., Athanasopoulos, P., Wiggett, A., Dering, B. and Kuipers, J. R.
2009. ‘Unconscious effects of language-specific terminology on pre-at-
tentive color perception’, Proceedings of the National Academy of Sciences USA
106, 4567–4570.
Thierry, G., Giraud, A.-L. and Price, C. 2003. ‘Hemispheric dissociation in
access to the human semantic system’, Neuron 38, 499–506.
Thierry, G., Martin, C., Gonzalez-Diaz, V., Rezzaie, R., Roberts, N. and
Davis, P. M. 2008. ‘Event-related potential characterisation of the
Shakespearean functional shift in narrative sentence structure’,
Neuroimage 40, 923–931.
Thierry, G. and Price, C. 2006. ‘Dissociating verbal and nonverbal concep-
tual processing in the human brain’. Journal of Cognitive Neuroscience 18,
1018–1028.
Thierry, G. and Wu, Y. J. 2007. ‘Brain potentials reveal unconscious transla-
tion during foreign language comprehension’, Proceedings of the National
Academy of Sciences, USA 104, 12530–12535.
Vaughan-Evans, A., Kuipers, J. R., Thierry, G. and Jones, M. 2014.
‘Anomalous transfer of syntax between languages’, The Journal of
Neuroscience 34, 8333–8335.
Wu, Y. J., Cristino, F., Leek, C. and Thierry, G. 2013. ‘Non-selective lexical
access in bilinguals is spontaneous and independent of input monitor-
ing: evidence from eye tracking’, Cognition 129, 418–425.
hear Chinese’, The Journal of Neuroscience 30, 7646–7651.
Wu, Y. J. and Thierry, G. 2012. ‘How reading in a second language protects
your heart’, The Journal of Neuroscience 32, 6485–6489.
Wu, Y. J. and Thierry, G. 2013. ‘Fast modulation of executive function
by language context in bilinguals’, The Journal of Neuroscience 33,
13533–13537.

26
Epilogue:
multi-competence and
the Translanguaging
Instinct
Li Wei
Despite the apparent connection, the concept of multi-competence (MC)

was not proposed as a corrective to the Chomskyan competence–perfor-
mance distinction (e.g. Chomsky 1965), nor was it intended to add to the
expanding list of usage-focused competences (e.g. communicative compe-
tence, symbolic competence). As discussed in detail in Chapter 1 of this
volume, MC was initially developed with the specific concern of the then
dominant monolingual perspective on second language acquisition (SLA),
which saw the monolingual native speaker as the standard-bearer, and
expected everything that the second language learner did to be measured
against the sole language of the monolingual native speaker. Since the
initial proposal, the concept of MC has captured the imagination amongst
not only SLA researchers but also a variety of language scientists and
professionals, and generated a large body of new knowledge that challenges
the conventional conceptualisations of language, the language learner and
language learning, as the chapters of this volume demonstrate. In this
epilogue I will first outline what I believe are the key contributions of the
concept of MC to language science generally, not just to SLA research, and
revisit the myth of monolingualism and its implications for linguistics
research in terms of design and methods. I will then outline my own idea
of the ‘Translanguaging Instinct’, which I believe is both in keeping with the
MC concept and will take it forward for the years to come.
26.1 The consequences of the MC perspective
In SLA research, the notion of language is often taken for granted and
equal to culturally conventionalised labels such as Albanian, Malay, or
Zulu. Yet, paradoxically it seems, a significant proportion of SLA

534 LI WEI
researchers adopt a theoretical approach to linguistics that sees lan-

guage as a representation system by human beings and a state of the
faculty of the human mind. The generative approach to SLA, for
example, asks to what extent second language learners can access
the so-called Universal Grammar to acquire the target language,
where the target language is postulated as a specific culturally con-
ventionalised set of surface structures, whilst UG is a mental repre-
sentation system which is hypothesised as a set of principles
governing all human languages. A critique of such an approach to
SLA is neither the intention of this chapter nor within its scope. But,
as Burton-Roberts (2004) points out, there is an inherent problem with
the logic of this line of enquiry. That is, if UG is supposed to be about
all languages, as Chomsky seemed to want it to be, then it cannot be
conceptualised as a natural, biological, genetic endowment, since
particular languages as we know them are historically evolved social
conventions; and if UG is about something entirely natural, biological
or genetic, then it cannot be a theory of the actual languages that
human beings learn and use in society.
26.1.1 On language
In my view, MC reconceptualises the concept of language by raising three
crucial questions:
(i) Is language a separate and discrete module in the human mind in

relation to other cognitive systems such as memory, attention, or
emotion?
(ii) If the multi-competent mind is a whole, where, if at all, does it divide
into different languages?
(iii) If the human mind does not divide different languages or between
language and other cognitive systems, should bilingualism and
multilingualism research be focused on how the language users
use the multiple linguistic and cognitive resources available in
social interaction, rather than which and how many languages
they know and use?
The MC concept urges us to rethink the ‘Modularity of Mind’ (MOM)

theory (Fodor 1983). According to the MOM theory, the human mind
consists of a series of innate neural structures, or modules, which are
encapsulated with distinctive information and for distinct functions.
Language is but one module of the human mind. Whilst theoretically
plausible, there is ample neuroanatomic evidence to suggest that separat-
ing language from the rest of the mind is pointless, as there is no
such thing as a language-only brain area. The brain areas that are
involved in processing language information are also involved in other
non-verbal processes. Furthermore, language processing cannot be wholly

Epilogue: multi-competence and the Translanguaging Instinct 535
independent of auditory and visual processes, just as cognitive processes

such as number processing and colour categorisation cannot be wholly
independent of language. In terms of multilingual language users, there is
increasing evidence that their language experience and cognitive capacity
are closely interconnected and mutually beneficial. Language, from the
MC perspective, then, is a multisensory and multimodal semiotic system
interconnected with other identifiable but inseparable cognitive systems.
In this sense, the MC view of language is compatible with that of social
semiotics but with a strong emphasis on the connections and interactions
with other human cognitive systems.
Similarly, it seems inconceivable from existing research evidence that
the human mind can be divided into different languages. Some earlier
experimental data did show that processing later acquired language
might involve certain neural networks that are not central to first lan-
guage processing. But that tells us more about the process of language
learning than about the representation of different languages in the
human mind. Likewise the findings that certain brain areas may be
involved in processing the lexical tones and logographic writing systems
that some of the world’s languages have point to the closer connections
between language and other cognitive systems as much as to the differ-
ences between languages. The MC perspective asks us to explore the
human mind as a holistic multi-competence, and not to count the num-
ber of languages.
The MC perspective also challenges the received and uncritical view in
some quarters of the applied and sociolinguistics communities that
bilingualism and multilingualism are about the protection of individual
languages and, since language and sociocultural identity are postulated
to be intrinsically linked, maintaining one’s language means maintain-
ing one’s identity. If the divide between languages and between language
and other cognitive systems has no psychological reality, shouldn’t the
sociocultural conventionalisation of different languages and the one-to-
one relationship between language and identity be reconsidered and
reconceptualised too? We know for a fact that the labelling of languages
is largely arbitrary and can be politically and ideologically charged.
There is often a close relationship between the identity of a language
and the nation state. But in everyday social interaction, multilingual
language users move freely and dynamically between the so-called lan-
guages, language varieties, styles, registers, and writing systems, in
order to fulfil a variety of strategic and communicative functions.
Code-switching, the alternation between languages, spoken, written or
signed, between language varieties, and between speech, writing
and signing, is a very common feature of human social interaction. It
constructs an identity for the speaker that is different from an La identity
or an Lb identity. Moreover, language users use semiotic resources
such as gesture or facial expression in conjunction with language to

536 LI WEI
communicate with each other. From an MC perspective, asking simply

which language is being used becomes an uninteresting and insignifi-
cant question. Language is a multisensory and multimodal resource that
human beings use in conjunction with other cognitive and semiotic
resources for thinking and for communicating thought.
26.1.2 On the language learner

Perhaps the most noticeable consequence of the MC perspective has been in
the reconceptualisation of the language learner as a legitimate, multi-com-
petent language user in their right and not ‘in relation to’ the so-called
monolingual native speaker. Even though the native speaker model in lan-
guage teaching and learning has been repeatedly critiqued in the applied and
sociolinguistics literature, not least by the proponents of the MC concept (e.g.
Cook 2007), in practice, the non-native language learner is often put in a
position where he or she is expected to produce linguistic forms that are
identical to those produced by native speakers. Their own linguistic and
sociocultural knowledge and resources tend to be regarded as irrelevant,
even counterproductive. By challenging the traditional views on what lan-
guage is, the MC perspective argues that, as long as one still uses language,
one is forever a language learner, because no one could ever grasp a language,
any language, in its entirety, nor will the knowledge of language in one’s
mind stay static. It therefore is more useful to think of language learners as
language users, whose linguistic knowledge will continue to develop and
change throughout the life-span (biological and psychological development
from birth to death) and life-course (events, transitions, and trajectories in
experiential and sociological terms).
26.1.3 On language learning

The process of learning a new language, whether first or second, native or
foreign, is always a multisensory and multimodal experience. It involves
complex mapping between sounds, objects, actions, and expressions and
draws on auditory and articulatory apparatus as well as memory and
attentional resources. It is also an acculturation process whereby the
language learner is socialised into a culturally specific way of interpreting
the intentional meaning and illocutionary force of human behaviour
including language use. The MC perspective emphasises the importance
of looking at language learning as a multisensory and multimodal as well
as acculturation process where the apparent failure to produce a target
linguistic structure may be caused by a variety of factors, most of which are
likely to be non-linguistic. It also opens up the possibility that language
learners, even at the very early stage of acquiring a new language, should
be able to create novel structures to effectively communicate meaning
and achieve understanding. Moreover, from a language socialisation

perspective, language learning is a multidirectional and multidimensional

adaptation process where the learner’s agency is equal to – if not more
important than – the learning model or target itself.
26.2 The myth of monolingualism
One of the central premises of MC is challenging the monolingual native

speaker norm that seems to dominate much of the thinking in SLA. It is
curious how a notion as ill-defined as monolingualism became so persis-
tent in an academic research field. It is true that some nation states have a
monolingual language policy and ideology, imposing one language on the
whole population despite the fact that there are different languages in
coexistence in the community. Such a policy and ideology usually runs
against the social reality of the nation state in question and only serves
political purposes, for example, to enforce an official or national language.
As far as individual language users are concerned, the notion of mono-
lingualism is simply a myth. A strict definition of monolingualism would
have to be that it is a condition where a person has only one language and
has no capacity to acquire more than one language. We know from ample
linguistic and cognitive science research that human beings are prewired
for language learning. At least in theory then, we have the capacity to learn
as many languages as we want to. If we were to find individuals with a
monolingual condition, we could legitimately ask what prevented them
from acquiring and using other languages. But equally important are
questions about that sole language of the so-called monolingual person.
Is it an accent, a set of words and sentences, a particular style or register, or
something else? To what level does one need to know that language in
order to be described as having acquired it? A few words and phrases, the
ability to have a simple and informal conversation, to deliver a formal
speech, or to read and write in the language?
As has been discussed above, the MC perspective urges us to rethink
language as one of the cognitive systems and semiotic resources that
human beings possess and use for communication. It includes what has
been culturally conventionalised as different languages including differ-
ent writing systems, as well as language varieties, styles, and registers. It is
also intricately linked to other human cognitive systems such as memory,
attention, and emotion. A multi-competent language user is one who
moves comfortably between registers, styles, varieties, and languages, as
well as across modalities – speech, writing, sign, gesture, etc. One can
measure the extent to which an individual language user moves between
and across linguistic codes and communication modes and ask why some
language users may do it more easily than others. But a true monolingual,
from this perspective, would have to be someone who simply does not
have the capacity to do any of this at all.

538 LI WEI
As to the proficiency level in a language, linguists of all shapes and forms

have struggled over the years to develop commonly acceptable measures.
Age of acquisition, frequency of use, structural salience, and functional
load have all been found to affect one’s language comprehension and
production. But the classification of advanced, intermediate, and beginner
language learner/user remains largely arbitrary. As has been said above,
the MC perspective sees the counting of the number of languages in the
human mind as pointless. By implication, it sees the measuring of profi-
ciency in different languages as largely uninteresting and misguided, as it
perpetuates the monolingual native speaker norm and the myth of
monolingualism.
The vast majority of the world’s population is born into families and
communities where multiple languages coexist. Unfortunately not all
languages are equally represented in education and the media. And as a
result, some people are socialised into using one language only most of
the time. But this does not mean that they have lost the capacity to
acquire and use other languages. It only raises issues of why not every
language gets equal institutional support and what is the process of
language attrition in cognitively healthy individuals. So, whilst not
everyone may be actively bilingual or multilingual in everyday life, a
true monolingual with no or very limited capacity to acquire and use
more than one language is a rarity.
Apart from the negative impact the myth of monolingualism has
had on society and in the ideological and policy realms, it has also
influenced the design of research on bilingual and multilingual lan-
guage users. It is still very common to see studies that compare
bilinguals and multilinguals on the one hand and the so-called mono-
linguals on the other. And often the latter – the monolinguals – are
used as the benchmark by which the bilingual’s and multilingual’s
behaviour is judged to be normal or deviant. The MC perspective
invites us to think about research design afresh, considering different
types of language users, all potentially multilingual and multimodal –
and asking a new set of questions in ways in which they have not
been asked before. These questions could include the following.
When, if ever, does a developing language user (i.e. someone who is
in the process of developing knowledge of a specific language and
language variety) become aware of the different labels, usage norms,
and structural differences, and what are the consequences of that
awareness, or the lack of it, on language processing? How does a
developing language user create new structures and convey meaning
with limited resources? How does a multi-competent language user
coordinate multiple semiotic resources and cognitive systems to con-
struct and convey meaning? How do the roles of different linguistic
and non-linguistic resources in language production and comprehen-
sion change over time across the life-span?

26.3 The Translanguaging Instinct
The idea that human beings are born with an innate capacity for
learning languages seems uncontroversial in very general terms. But
the popularisation of the idea through the metaphor of the ‘Language
Instinct’ (Pinker 1994) has caused a considerable amount of debate
both in the scientific community and amongst the general public
over the nature of that innate capacity. Part of the controversy is
caused by the use of the metaphor ‘instinct’, which suggests a fixed
pattern of behaviour in animals triggered by some external stimuli. If
language were an instinct in this sense, it would run against
the whole argument that Chomsky put forward for the existence of
UG against the behaviourist approaches to language acquisition.
Nevertheless, some psychologists and psycholinguists have continued
with the instinct metaphor, and argued that child language acquisi-
tion is the result of what they call the ‘Interactional Instinct’, a
biologically based drive for infants and children to attach, bond,
and affiliate with conspecifics in an attempt to become like them
(Lee et al. 2009; Joaquin and Schumann 2013). This biologically based
drive provides neural structures that entrain children acquiring their
languages to the faces, voices, and body movements of caregivers.
The drive, or the Interactional Instinct, also determines the relative
success of older adolescents and adults in learning additional lan-
guages later in life due to the variability of individual aptitude and
motivation as well as environmental conditions. The notion of lan-
guage in the Interactional Instinct framework is very similar to that
from the MC perspective in incorporating multiple semiotic and
cognitive systems.
Many linguists have argued that interaction – mutual activity which
requires the involvement of at least two persons and which causes a
mutual effect – is the foundation of human sociality (Enfield and
Levinson 2006). In the twenty-first century, much of human social
interaction is mediated through multimedia technology. The salience
of mediated interaction in everyday life helps to remind us of the
multisensory and multimodal process of language learning and lan-
guage use. It is within this context that the idea of the
Translanguaging Instinct has been developed (Li Wei 2011; Garcia and
Li Wei 2014).
The notion of translanguaging is not some fancy post-modernist term
to replace traditional terms such as code-switching or language crossing
in referring to specific language mixing behaviour. Translanguaging is
the dynamic process whereby multilingual language users mediate com-
plex social and cognitive activities through strategic employment of
multiple semiotic resources to act, to know, and to be (Garcia and Li

540 LI WEI
Wei 2014). In line with the MC perspective, translanguaging emphasises

the interconnectedness between the traditionally and conventionally
understood languages and other human communication systems. In
our view, human beings’ knowledge of language cannot be separated
from their knowledge of human relations and human social interaction,
which includes the history, the context of usage, and the emotional and
symbolic values of specific socially constructed languages. The trans
prefix in translanguaging emphasises:
– the fluid practices that go beyond, i.e. transcend, socially constructed

language systems and structures to engage diverse multiple meaning-
making systems and subjectivities;
– the transformative capacity of the translanguaging process, not only for
language systems but also the individual’s cognition and social
structures;
– the transdisciplinary consequences of re-conceptualising language, lan-
guage learning, and language use for linguistics, psychology, sociology,
and education.
Moreover, translanguaging enables creativity and criticality in the lan-

guage user that Li Wei (2011) defined as
the ability to choose between following and flouting the rules and norms
of behaviour, including the use of language, and to push and break
boundaries between the old and the new, the conventional and the origi-
nal, and the acceptable and the challenging
and
the ability to use available evidence appropriately, systematically and

insightfully to inform considered views of cultural, social and linguistic
phenomena, to question and problematise received wisdom, and to
express views adequately through reasoned responses to situations.
(Li Wei 2011, p. 1224)
These two concepts are intrinsically linked: one cannot push or break
boundaries without being critical; and the best expression of one’s criti-
cality is one’s creativity. Multilingualism, by the very nature of the
phenomenon, is a rich source of creativity and criticality, as it entails
tension, conflict, competition, difference, and change in a number of
spheres, ranging from attentional control and emotions to histories and
ideologies.
Translanguaging emphasises the multisensory, multimodal, multi-
semiotic, and multilingual nature of the meaning-making process. In
this process a ‘principle of abundance’ is in operation, that is, as
many different cues should be produced simultaneously as possible
in order to facilitate communication. For example, everyday commu-
nication at a very basic level involves traditional linguistic signs

(letters), sounds, images, emoticons, and pictures. Nicholas and Starks

(2014) use the heart image that has become an iconic element in
expressions of affection for particular cities, for example, ‘I (heart
sign in red) NY’, to illustrate the interconnectedness of all signs. It
can also be seen as a rudimentary example of the process of trans-
languaging, where an image of a heart (traditionally understood as a
noun) is used in the place in the linguistic construction that is usually
occupied by a verb. Yet when the expression is ‘read out’, most
people would say ‘I love New York’ rather than ‘I heart New York’,
so changing the grammatical status of the word to which the image is
linked, unless it is deliberately nominalised for some humorous effect
as in Lindsey Kelk’s paperbacks. As Nicholas and Starks argue, ‘exam-
ples such as these reinforce the variation and creativity of speakers as
they bring together multiple elements of rich and complex commu-
nicative resources’. Research evidence shows that children, even
infants, have no problem using their multiple semiotic resources
to interpret different forms of symbolic references (Namy and
Waxman 1998; Plester et al. 2011). Human beings have a natural
Translanguaging Instinct, an innate capacity to draw on as many
different cognitive and semiotic resources as are available to them
to interpret meaning intentions and to design actions accordingly.
This innate capacity drives humans to go beyond narrowly defined
linguistic cues and transcend the culturally defined language bound-
aries to achieve effective communication.
Whilst the natural tendency to combine all available cognitive,
semiotic, sensory, and modality resources in language learning and
language use is innate, one relies on different resources differentially
across one’s life-span and life-course, as not all resources are equally
available at all times. In first language acquisition, infants naturally
draw meaning from a combination of sound, image, and action, and
the sound-meaning mapping in word learning crucially involves
image and action. The resources needed for literacy acquisition are
called upon later. In bilingual first language acquisition where cross-
linguistic equivalents are learned, the child additionally learns to
associate the target word with a specific context or addressee as
well as contexts and addressees where either language is acceptable,
giving rise to the possibility of code-switching. In second language
acquisition in adolescence and adulthood, some resources become
less available, for example, resources required for tonal discrimina-
tion, while others can be enhanced by experience and become more
salient in language learning and use, for example, resources required
for analysing and comparing syntactic structures and pragmatic func-
tions of specific expressions. As people become more involved in
complex communicative tasks and demanding environments, the
natural tendency to combine multiple resources drives them to

542 LI WEI
look for more cues and exploit different resources. They will also
learn to use different resources for different purposes, resulting in
functional differentiation of different linguistic resources (e.g. accent,
writing) and between linguistic and other cognitive and semiotic
resources. Crucially, the innate capacity to exploit multiple resources
will not be diminished over time; in fact it is enhanced with experi-
ence. Critical analytic skills are developed in terms of understanding
the relationship between the parts (specific sets of skills, such as
counting, drawing, singing) and the whole (multi-competence and
the capacity for coordination between the skill subsets) in order to
functionally differentiate the different resources required for differ-
ent tasks.
One consequence of the translanguaging perspective on bilingualism
and multilingualism research is making the comparison between L1
and L2 acquisition – purely in terms of attainment – insignificant.
Instead, questions should be asked as to what resources are needed,
available, and being exploited for specific learning tasks throughout
the life-span and life-course. Why are some resources not available at
certain times, and what do language users do when some resources
become difficult to access? How do language users combine the avail-
able resources differentially for specific tasks? In seeking answers to
these questions, the multisensory, multimodal, and multilingual nature
of human learning and interaction is at the centre of the
Translanguaging Instinct hypothesis and that of the concept of multi-
competence.
References
Burton-Roberts, N. 2004. ‘Linguistic diversity in a universalist context:

language, thought and culture’, in R. Talif, S. H. Chan, B. E. Wong,
A. N. Abdullah and R. Noor (eds), Beyond Barriers, Fresh Frontiers: Selected
Readings on Languages, Literatures, and Cultures, pp. 112–124. Serdang
Selangor: Universiti Putra Malaysia Press.
Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.
Cook, V. J. 2007. ‘The nature of the L2 user’, in L. Roberts, A. Gurel, S. Tatar
and L. Marti (eds), EUROSLA Yearbook 7, 205–220.
Enfield, N. J. and Levinson, S. C. 2006. Roots of Human Sociality: Culture,
Cognition and Interaction. Oxford: Berg.
Fodor, J. A. 1983. Modularity of Mind: An Essay on Faculty Psychology.
Education. Houndsmill, Basingstoke: Palgrave Macmillan.
Joaquin, A. D. L. and Schumann, J. H. (eds) 2013. Exploring the Interactional
Instinct. Oxford: Oxford University Press.

Lee, N., Mikesee, L., Joaquin, A. D. L., Mates, A. W. and Schumann, J. H.

2009. The Interactional Instinct: The Evolution and Acquisition of Language.
Oxford: Oxford University Press.
of Pragmatics 43 (5), 1222–1235.
Namy, L. L. and Waxman, S. R. 1998. ‘Words and gestures: infants’ inter-
pretations of different forms of symbolic reference’, Child Development
69 (2), 295–308.
Nicholas, H. and Starks, D. 2014. Language Education and Applied Linguistics:
Bridging the Two Fields. Abingdon, Oxfordshire: Routledge.
Pinker, S. 1994. The Language Instinct: The New Science of Language and Mind.
New York: William Morrow and Co.
Plester, B., Lerkkanen, M.-K., Linjama, L. J., Rasku-Puttonen, H. and
Littleton, K. 2011. ‘Finnish and UK English pre-teen children’s text mes-
sage language and its relationship with their literacy skills’, Journal of
Computer Assisted Learning 27 (1), 37–48.

Bibliography of
‘multi-competence’
(and related topics)
Compiled by Goro Murahata, Yoshiko Murahata
and Vivian Cook
This bibliography brings together books and papers that either explicitly
mention multi-competence or are closely related to the multi-competence
perspective.
Agrillo, C. and Roberson, D. 2008. ‘Colour language and colour cognition:

Brown and Lenneberg revisited’, Visual Cognition 17, 412–430.
Akamatsu, N. 2010. Nihonjin Eigo Gakushusha no Kanshi Shiyo to Ninchi
[The use of the English articles by Japanese learners of English and their
cognition]. Paper read at the 36th Annual Conference of The Society of
English Language Education, Kansai University, Osaka, Japan.
Akamatsu, N. 2011. ‘Does a foreign language affect cognition?
Re-evaluating evidence for L2 effects on bilinguals’ object classification’,
Studies in English Language Teaching (The Kansai English Language
Education Society) 34, 1–10.
Akbar, F. S. 2013. ‘The case against Monolingual Bias in Multilingualism’,
Teachers College, Columbia University Working Papers in TESOL & Applied
Linguistics 13 (2), 42–44.
Alotaib, H. 2014. ‘The validity of likeability as a new construct of ESL
pronunciation’, Advances in Language and Literary Studies 5, 48–58.
Alptekin, C. 2010. ‘Redefining multi-competence for bilingualism and
EFL’, International Journal of Applied Linguistics 20, 95–110.
Andrews, D. 1994. ‘The Russian color categories sinij and goluboj: an
experimental analysis of their interpretation in the standard and émigré
languages’, Journal of Slavic Linguistics 2, 9–28.
Benjamins.

Bibliography of ‘multi-competence’ (and related topics) 545
Athanasopoulos, P. 2006. ‘Effects of grammatical representation of

9, 89–96.
Athanasopoulos, P. 2009. ‘Cognitive representation of color in bilin-
guals: the case of Greek blues’, Bilingualism: Language and Cognition
12, 83–95.
Athanasopoulos, P. 2011a. ‘Color and bilingual cognition’, in V. J. Cook and
B. Bassetti (eds), pp. 241–261.
Athanasopoulos, P. 2011b. ‘Cognitive restructuring in bilingualism’, in
A. Pavlenko (ed.), pp. 29–65.
Athanasopoulos, P. and Aveledo, R. 2012. ‘Linguistic relativity and bilin-
gualism’, in J. Altarriba and L. Isurin (eds), Memory, Language, and
Bilingualism: Theoretical and Applied Approaches, pp. 236–255. Cambridge
University Press.
Japanese blues’, Bilingualism: Language and Cognition 14 (special issue),
9–17.
Athanasopoulos, P., Sasaki, M. and Cook, V. J. 2004. Do bilinguals think
differently from monolinguals? Evidence from colour categorization by
speakers of Japanese. Paper presented at the 14th European Second
Language Association Conference, San Sebastian, Spain.
Azuma, S. 2000. Bailingalizumu [Bilingualism]. Tokyo: Kodansha.
Bak, T. H. and Alladi, S. 2014. ‘Can being bilingual affect the onset of
dementia?’, Future Neurology 9 (2), 101–103.
Balcom, P. 1995. ‘Argument structure and multi-competence’, Linguistica
Atlantica (Memorial University of Newfoundland) 17, 1–17.
Balcom, P. 1999. ‘These constructions don’t acquire easily: middle con-
structions and multi-competence’, Canadian Journal of Applied Linguistics 2,
5–20.
Balcom, P. 2003. ‘Cross-linguistic influence of L2 English on Middle
Constructions in L1 French’, in V. J. Cook (ed.), pp. 168–192.
Bassetti, B. 2007. ‘Bilingualism and thought: grammatical gender and
concepts of objects in Italian–German bilingual children’, International
Journal of Bilingualism 11, 251–273.
Bassetti, B. 2011. ‘The grammatical and conceptual gender of animals in
second language users’, in V. J. Cook and B. Bassetti (eds), pp. 357–384.
Bassetti, B. 2014. ‘Is grammatical gender considered arbitrary or semanti-
cally motivated? Evidence from young adult monolinguals, second

546 Goro Murahata, Yoshiko Murahata and Vivian Cook
language learners and early bilinguals’, British Journal of Psychology 105 (2),
273–297.
Bialystok, E., Craik, F. I. M. and Freedman, M. 2007. ‘Bilingualism as
a protection against the onset of symptoms of dementia’,
Birdsong. D. 2005. ‘Nativelikeness and non-nativelikeness in L2A
Block, D. 2007. Second Language Identities. London: Continuum.
politeness in family discourse’, Journal of Pragmatics 14 (2), 259–288.
Boroditsky, L., Ham, W. and Ramscar, M. 2002. ‘What is universal in event
perception? Comparing English & Indonesian speakers’, in W. D. Gray
and C. D. Schunn (eds), Proceedings of the 24th Annual Conference of the
Cognitive Science Society, pp. 136–141. Lawrence Erlbaum.
Boroditsky, L., Schmidt, L. and Phillips, W. 2003. ‘Sex, syntax, and seman-
tics’, in D. Gentner and S. Goldin-Meadow (eds), Language in Mind:
Advances in the Study of Language and Cognition, pp. 61–79. Cambridge,
MA: MIT Press.
Braine, G. 2010. Nonnative Speaker English Teachers: Research, Pedagogy, and
Professional Growth. New York and London: Routledge.
Brown, A. 2007. Crosslinguistic Influence in First and Second Languages:
Convergence in Speech and Gesture. Unpublished Doctoral Dissertation,
Boston University.
Language Assessment Quarterly 10, 219–235.
Brown, A. and Gullberg, M. 2008. ‘Bidirectional crosslinguistic influ-
ence in L1–L2 encoding of manner in speech and gesture: a study of
Japanese speakers of English’, Studies in Second Language Acquisition 30,
225–251.
in event conceptualization? Expressions of Path among Japanese
learners of English’, Bilingualism: Language and Cognition 14, 79–94.
variation in clausal packaging in Japanese’, Second Language Research
28 (4), 415–442.
Brown, A. and Gullberg, M. 2013. ‘L1–L2 convergence in clausal packaging
in Japanese and English’, Bilingualism: Language and Cognition 16, 477–494.
Brown, R. W. 1957. ‘Linguistic determinism and the part of speech’, Journal
of Abnormal and Social Psychology 55, 1–5.
Brown, R. W. and Lenneberg, E. H. 1954. ‘A study in language and cogni-
tion’, Journal of Abnormal and Social Psychology 49, 454–462.


Bylund, E. 2011. ‘Language-specific patterns in event conceptualization:
insights from bilingualism’, in A. Pavlenko (ed.), pp. 108–142.
Bylund, E. and Athanasopoulos, P. (eds) 2015. Cognitive SLA: The Language
and Thought of Motion in Second Language Speakers, special issue of Modern
Language Journal 99.
Chang, C. B. 2011. ‘Systemic drift of L1 vowels in novice L2 learners’, in
W.-S. Lee and E. Zee (eds), Proceedings of the 17th International Congress of
Phonetic Sciences, pp. 428–431. City University of Hong Kong.
249–268.
Chang, C. B. and Mishler, A. 2012. ‘Evidence for language transfer leading
to a perceptual advantage for non-native listeners’, Journal of the Acoustical
Society of America 132 (4), 2700–2710.
Choong, K. P. 2006. ‘Multi-competence and second language teaching’,
Teachers College, Columbia University Working Papers in TESOL & Applied
Linguistics 6, 1. www.tc.columbia.edu/academic/tesol/WJFiles/pdf/
Choong.pdf
Coggins, P. E., Kennedy, T. J. and Armstrong, T. A. 2004. ‘Bilingual corpus
callosum variability’, Brain and Language 89 (1), 69–75.
557–591. Reprinted in L. Ortega (ed.) 2010, Second-Language Acquisition:
Critical Concepts, vol. I: Foundations of Second Language Acquisition.
Routledge.
Cook, V. J. 1995a. ‘Multi-competence and the learning of many languages’,
Language, Culture and Curriculum 8, 93–98. Reprinted in K. de Bot, W. Lowie
and M. Verspoor (eds) 2005, Second Language Acquisition: An Advanced
Resource Book. Abingdon: Routledge.
Cook, V. J. 1995b. ‘Multi-competence and effects of age’, in D. Singleton
and Z. Lengyel (eds), The Age Factor in Second Language Acquisition: A Critical
Look at the Critical Period Hypothesis, pp. 51–66. Clevedon, UK: Multilingual
Matters.
Cook, V. J. 1996. ‘Competence and multi-competence’, in G. Brown,
K. Malmkjaer and J. Williams (eds), Performance and Competence in Second
Language Acquisition, pp. 57–69. Cambridge University Press.
Cook, V. J. 1999. ‘Going beyond the native speaker in language teaching’.
TESOL Quarterly 33, 185–209.

Cook, V. J. 2001a. ‘Linguistics and second language acquisition: one person

with two languages’, in M. Aronoff and J. Rees-Miller (eds), The Handbook
of Linguistics, pp. 488–511. Malden, MA: Blackwell.
Cook, V. J. 2001b. ‘Using the first language in the classroom’, Canadian
Modern Language Review 57, 402–423.
Cook, V. J. (ed.) 2002a. Portraits of the L2 User. Clevedon, UK: Multilingual
Matters.
Cook, V. J. 2002b. ‘Background to the L2 user’, in V. J. Cook (ed.),
pp. 1–28.
Cook, V. J. 2002c. ‘Language teaching methodology and the L2 user
perspective’, in V. J. Cook (ed.), pp. 327–343.
Cook, V. J. (ed.) 2003a. Effects of the Second Language on the First. Clevedon,
Cook, V. J. 2003b. ‘The changing L1 in the L2 user’s mind’, in V. J. Cook (ed.),
pp. 1–18.
Cook, V. J. 2006. ‘Interlanguage, multi-competence and the problem of
the “second” language’, Rivista di Psicolinguistica Applicata 6, 39–52.
Cook, V. J. 2007a. ‘The nature of the L2 user’, EUROSLA Yearbook 7, 205–220.
Reprinted in Li Wei (ed.) 2011, The Routledge Applied Linguistics Reader,
pp. 77–89. Routledge.
Cook, V. J. 2007b. ‘The goals of ELT: reproducing native-speakers or
promoting multi-competence among second language users?’, in
J. Cummins and C. Davidson (eds), International Handbook of English
Language Teaching, pp. 237–248. Boston, MA: Kluwer.
Cook, V. J. 2007c. ‘Multi-competence: black hole or wormhole for SLA
research?’, in Z.-H. Han (ed.), Understanding Second Language Process,
pp. 16–26. Clevedon: Multilingual Matters.
Cook, V. J. 2008a. ‘Linguistic contributions to bilingualism’, in J. Altarriba
and R. R. Heredia (eds), An Introduction to Bilingualism: Principles and
Processes, pp. 245–264. New York: Lawrence Erlbaum Associates.
Cook, V. J. 2008b. Second Language Learning and Language Teaching, fourth
edition. London: Hodder Education.
Cook, V. J. 2009a. ‘Language user groups and language teaching’, in
Cook, V. J. 2009b. ‘Multilingual Universal Grammar as the norm’, in
I. Leung (ed.), Third Language Acquisition and Universal Grammar,
Cook, V. J. 2009c. ‘The multi-competence perspective and the teaching
of British Sign Language’, in M. Mertanzi (ed.), Sign Language Teaching
and Learning, pp. 135–143. Centre for Deaf Studies, University of
Bristol.
Applied Linguistics, pp. 6–22. London: Palgrave Macmillan.


of Applied Linguistics, pp. 3768–3774. Oxford, UK: Wiley-Blackwell.
Cook, V. J. 2015. ‘Discussing the language and thought of motion in
second language speakers’, in E. Bylund and P. Athanasopoulos (eds),
pp. 154–164.
Cook, V. J., to appear. ‘The relationship between the languages of
multi-competence’, in R. Alonzo (ed.), Cross-linguistic Influence. Bristol:
Cook, V. J., Alenazi, O., Alkadi, H. and Almutrafi, F., in progress. ‘Bilingual
street signs and eyetracking’.
Cook, V. J. and Bassetti, B. (eds) 2011. Language and Bilingual Cognition. Hove,
UK: Psychology Press.
Japanese L2 users of English’, International Journal of Bilingualism 10,
137–152.
Cook, V. J., Iarossi, E., Stellakis, N. and Tokumaru, Y. 2003. ‘Effects of
the L2 on the syntactic processing of the L1’, in V. J. Cook (ed.),
pp. 192–213.
Costa, A. Foucart, A., Arnon, Aparici, M. and Apesteguia, J. 2014. ‘ “Piensa”
twice: on the foreign language effect in decision making’, Cognition 130,
236–254.
Costa, A., Foucart, A., Hayakawa, S., Aparici, M., Apesteguia, J., Heafner, J.
and Keysar, B. 2014. ‘Your morals depend on language’, PLoS ONE 9 (4),
e94842.
Coventry, K. R., Guijarro-Fuentes, P. and Valdes, B. 2011. ‘Spatial language
and second language acquisition’, in V. J. Cook and B. Bassetti (eds),
pp. 263–286.
Davies, A. 2003. The Native Speaker: Myth and Reality. Clevedon: Multilingual
Matters.
De Angelis, G. and Dewaele, J.-M. 2011. New Trends in Crosslinguistic Influence
and Multilingualism Research. Clevedon: Multilingual Matters.
De Groot, A. M. B. 2002. ‘Lexical representation and lexical processing
in the L2 user’, in V. J. Cook (ed.), pp. 32–63.
An Introduction. Hove, UK: Psychology Press.

Della Rosa, P. A., Videsott, G., Borsaa, V. M., Canini, M., Weekes, B. S.,
Franceschini, R. and Abutalebi, J. 2013. ‘A neural interactive location
for multilingual talent’, Cortex 49 (2), 605–698.
Dewaele, J.-M. 2010. Emotions in Multiple Languages. Basingstoke: Palgrave
Macmillan.
Dewaele, J.-M. and Costa, B. 2013. ‘Multilingual clients’ experience of
psychotherapy’, Language and Psychoanalysis 2 (2), 31–50.
competence’, International Journal of Multilingualism 9, 352–366.
tolerance of ambiguity?’, Bilingualism: Language and Cognition 16 (1),
231–240.
guals’, International Journal of Bilingualism 18 (3), 203–221.
Djigunovic, J. M. 2010. ‘Starting age and L1 and L2 interaction’, The
International Journal of Bilingualism 14 (3), 303–314.
Dussias, P. E. and Sagarra, N. 2007. ‘The effect of exposure on syntactic
Cognition 10 (1), 101–116.
Edwards, M. and Dewaele J.-M. 2007. ‘Trilingual conversations: a win-
dow into multi-competence’, International Journal of Bilingualism 11,
221–242.
Ellis, E. 2006. ‘Monolingualism: the unmarked case’, Estudios de
Sociolingüı́stica 7 (2), 173–196.
Ervin-Tripp, S. 2011. ‘Advances in the study of bilingualism: a personal
view’, in V. J. Cook and B. Bassetti (eds), pp. 219–228.
Ewert, A. 2008. ‘L1 syntactic preferences of Polish adolescents in bilingual
and monolingual education programmes’, in D. Gabrys-Barker (ed.),
Morphosyntactic Issues in Second Language Acquisition, pp. 47–62. Clevedon,
Ewert, A. 2010. ‘An educational language community: external and
internal language use by multilingual students’, in J. Arabski and
A. Wojtaszek (eds), Neurolinguistic and Psycholinguistic Perspective on SLA,
pp. 159–174. Clevedon, UK: Multilingual Matters.
Ewert, A., forthcoming. L2 Users’ L1. Bristol: Multilingual Matters.
Filippi, R., Morris, J., Richardson, F. M., Bright, P., Thomas, M. S. C.,
Karmiloff-Smith, A. and Marian, V. 2015. ‘Bilingual children show an
advantage in controlling verbal interference during spoken language
comprehension’, Bilingualism: Language and Cognition 18 (3), 490–501.
Forbes, J. N., Poulin-Dubois, D., Rivero, M. R. and Sera, M. 2008.
‘Grammatical gender affects bilinguals’ conceptual gender: implications

for linguistic relativity and decision making’, The Open Applied Linguistics
Journal 1, 68–76.
Friesen, D. C. and Jared, D. 2011. ‘Cross-language phonological activation
of meaning: evidence from category verification’, Bilingualism: Language
Green, D. W. 1998. ‘Mental control of the bilingual lexico-semantic
system’, Bilingualism: Language and Cognition 1, 67–81.
Green, D. W. 2011. ‘Bilingual worlds’, in V. J. Cook and B. Bassetti (eds),
229–240.
Grosjean, F. 1985. ‘The bilingual as a competent but specific speaker-
hearer’, Journal of Multilingual and Multicultural Development 6 (6), 467–477.
Grosjean, F. 1989. ‘Neurolinguists, beware! The bilingual is not two
Grosjean, F. 2008. Studying Bilinguals. Oxford: Oxford University Press.
Gullberg, M. 2011. ‘Thinking, speaking and gesturing about motion in
more than one language’, in A. Pavlenko (ed.), pp. 143–169.
Hall, J. and Navarro, N. 2011. ‘Lessons for WAC/WID from language learn-
ing research: multi-competence, register acquisition, and the college
writing student’, Across the Disciplines 8, 4. http://wac.colostate.edu/atd/
ell/hall-navarro.cfm
Hall, J. K., Cheng, A. and Carlson, M. T. 2006. ‘Reconceptualizing multi-
220–240.
Halsband, U., Krause, B. J., Sipiläd, H., Teräs, M. and Laihinen, A. 2002. ‘PET
studies on the memory processing of word pairs in bilingual Finnish–
English subjects’, Behavioural Brain Research 132 (1), 47–57.
Han, A. Y. and Ginsburg, H. P. 2001. ‘Chinese and English mathematics
language: the relation between linguistic clarity and mathematics
performance’, Mathematical Thinking and Learning 3, 201–220.
Han, Z.-H. and Cadierno, T. (eds) 2010. Linguistic Relativity in SLA: Thinking
for Speaking. Bristol: Multilingual Matters.
Harada, T. 2007. ‘The production of voice onset time (VOT) by English-
speaking children in a Japanese immersion program’, IRAL 45,
353–378.
He, A. W. 2013. ‘The Wor(l)d is a collage: multi-performance by Chinese
heritage language speakers’, Modern Language Journal 97 (2), 304–317.
Heinz, B. 2003. ‘Backchannel responses as strategies responses in bilingual
speakers’ conversations’, Journal of Pragmatics 35, 1113–1142.
Hendriks, H. and Hickmann, M. 2011. ‘Expressing voluntary motion in a
second language: English learners of French’, in V. J. Cook and B. Bassetti
(eds), pp. 315–339.


Herlihy-Mera, J. 2013. ‘Reinterpreting “Papa(á) in Cuba: on the social
dimensions of Hemingway’s translingual nickname’, The Hemingway
Review 33 (1), 100–109.
Hermans, D., Ormel, E., van Besselaar, R. and van Hell, J. 2011. ‘Lexical
activation in bilinguals’ speech production is dynamic: how language
ambiguous words can affect cross-language activation’, Language and
Cognitive Processes 26 (10), 1687–1709.
Hirata-Edds, T. 2011. ‘Influence of second language Cherokee immersion
on children’s development of past tense in their first language, English’,
Language Learning 61 (3), 700–733.
Hunt, E. and Agnoli, F. 1991. ‘The Whorfian hypothesis: a cognitive
psychological perspective’, Psychological Review 98, 377–389.
Imai, M. and Gentner, D. 1997. ‘A cross-linguistic study of early word
169–200.
Ishikawa, K. 2005. Kotoba To Shinri [Language and mind]. Tokyo: Kuroshio
Publishing Company.
Ismail, A. 2012. ‘An evaluation of the monolingual fallacy in Oman’,
International Journal of Applied Linguistics 22 (2), 143–159.
Jarvis, S. 2003. ‘Probing the effects of the L2 on the L1: a case study’, in
V. J. Cook (ed.), pp. 81–102.
Jessner, U. 2003. ‘A dynamic approach to language attrition in multilingual
systems’, in V. J. Cook (ed.), pp. 234–246.
Jessner, U. 2008. ‘Multi-competence approaches to language proficiency
development in multilingual education’, in J. Cummins and
N. H. Hornberger (eds), Encyclopedia of Language and Education, second
edition, pp. 1552–1565. Berlin, Germany: Springer Science and
Business Media LLC.
Ji, L.-J., Zhang, Z. and Nisbett, R. E. 2004. ‘Is it culture or is it language?
Examination of language effects in cross-cultural research on categor-
ization’, Journal of Personality and Social Psychology 87 (1), 57–65.
Kasai, C. and Takahashi, J. A. 2005. ‘Effects of L2 on the cognitive
decision of shapes and mass: a report of the cases of Japanese univer-
sity students’, Bulletin of the Faculty of Regional Studies (Gifu University)
17, 21–33.
Kasai, C., Takahashi, J. A., Hiramoto, M. and Yamato, R. 2006. ‘Bilingual
cognition: effects of L2 on the cognitive decision of shapes and mass. A
report of the cases of Japanese bilinguals in Japan and in the UK’, Bulletin
of the Faculty of Regional Studies (Gifu University) 18, 91–102.
Kecskes, I. 1998. ‘The state of L1 knowledge in foreign language learners’,
Word 49, 321–340.

Kecskes, I. 2010. ‘Dual and multilanguage systems’, International Journal of

Multilingualism 7 (2), 91–109.
Kecskes, I. and Papp, T. 2000. Foreign Language and Mother Tongue. Mahwah,
Kecskes, I. and Papp, T. 2003. ‘How to demonstrate the conceptual effect of
L2 on L1?’, in V. J. Cook (ed.), pp. 247–265.
Klein, W. and Perdue, C. 1997. ‘The basic variety (or: couldn’t natural
languages be much simpler?)’, Second Language Research 13 (4),
301–347.
Kobayashi, H. and Rinnert, C. 2013. ‘L1/L2/L3 writing development: longi-
tudinal case study of a Japanese multi-competent writer’, Journal of Second
Language Writing 22, 4–33.
Kramsch, C. 2012. ‘Authenticity and legitimacy in multilingual SLA’,
Critical Multilingualism Studies 1 (1), 107–128.
Krause-Ono, M. 2004. ‘Change in backchannelling behaviour: the influ-
ence from L2 to L1 on the use of backchannel’, Studies of Cognitive Science
(Society of Muroran Cognitive Science) 3, 51–81.
Kuo, J. Y. and Sera, M. D. 2009. ‘Classifier effects on human categorization:
the role of shape classifiers in Mandarin Chinese’, Journal of East Asian
Kurinski, E. and Sera, M. D. 2011. ‘Does learning Spanish grammatical
gender change English-speaking adults’ categorization of inanimate
objects?’, Bilingualism: Language and Cognition 14, 203–220.
Kweon, S.-O. and Bley-Vroman, R. 2011. ‘Acquisition of the constraints on
wanna contraction by advanced second language learners: Universal
Grammar and imperfect knowledge’, Second Language Research 27 (2),
207–228.
Kwok, V., Niu, Z., Kay, P., Zhou, K., Jin, Z., So, K.-F. and Tan, L.-H. 2011.
‘Learning new color names produces rapid increase in gray matter in the
intact adult human cortex’, PNAS 108 (16), 6686–6688.
Laufer, B. 2003. ‘The influence of L2 on L1 collocational knowledge and
on L1 lexical diversity in free written expression’, in V. J. Cook (ed.),
pp. 19–31.
Leikin, M. 2012. ‘The effect of bilingualism on creativity: developmental
and educational perspectives’, International Journal of Bilingualism 17 (4),
431–447.
Li Wei 2011. ‘Multilinguality, multimodality and multi-competence: code-
and modeswitching by minority ethnic children in complementary
schools’, Modern Language Journal 95, 370–384.
Liu, Y. and Carney, W. 2012. ‘Multi-competence at the discourse level: a
comparison of persuasive essays by Chinese college and high school EFL
students’, Journal of Language Teaching and Research 3, 1061–1070.
Ljungberg, J. K., Hansson, P., Andres, P., Josefsson, M. and Nilsson, L.-G.
2013. ‘A longitudinal study of memory advantages in bilinguals’, PLoS
One 8 (9), e73029.

Lowe, E. W.-J. 2006. ‘Multi-competence and first language attrition:

where do we draw the line?’, Teachers College, Columbia University Working
Papers in TESOL & Applied Linguistics 6, 1. www.tc.columbia.edu/academic/
tesol/WJFiles/pdf/Lowe.pdf
Lucy, J. 2011. ‘Language and cognition: the view from anthropology’, in
V. J. Cook and B. Bassetti (eds), pp. 43–68.
MacKenzie, I. 2012. ‘English as a lingua franca in Europe: bilingualism
and multi-competence’, International Journal of Multilingualism 9,
83–100.
Malt, B. C., Li, P., Pavlenko, A. and Ameel, E. 2015. ‘Bilingual lexical inter-
action in late immersed Mandarin–English bilinguals’, Journal of Memory
Mennen, I. 2004. ‘Bi-directional interference in the intonation of Dutch
speakers of Greek’, Journal of Phonetics 32, 543–563.
Mitchell, R., DalNegro, S. and Ioannidou, E. 2012. ‘European multilingual-
ism, “multi-competence” and foreign language education’, in P. Studer
and I. Werlen (eds), Linguistic Diversity in Europe: Current Trends and
Discourses, pp. 203–232. Berlin, Germany: De Gruyter Mouton.
Murahata, G. 2010a. The influence of FLES on children’s cognition. Paper
read at the 36th Annual Conference of The Society of English Language
Education, Kansai University, Osaka, Japan.
Murahata, G. 2010b. Multi-cognition: evidence from a picture categoriza-
tion task by incipient child L2 users. Paper read at the 20th Annual
Conference of European Second Language Association, Reggio Emilia,
Italy.
Murahata, G. 2010c. ‘Conceptual preference modification in incipient
Japanese–English bilingual children’, Bulletin of Shikoku Society of English
Language Education 30, 55–64.
Murahata, G. 2010d. ‘Multi-cognition in L2 users: more evidence from an
object categorization task by Japanese elementary school children’,
Research Reports of Kochi University 59, 131–146.
Murahata, G. 2011. ‘The impact of L2 learning on children’s cognition:
evidence from an individuated object–substance distinction task’,
Annual Review of English Language Education in Japan (The Japan Society of
English Language Education) 22, 49–62.
Murahata, G. and Murahata, Y. 2007. ‘Does L2 learning influence cogni-
tion? Evidence from categorical and thematic organization of objects’,
Research Reports of the Department of International Studies (The Department
of International Studies, Faculty of Humanities and Economics, Kochi
University) 8, 17–27.
Murahata, G. and Murahata, Y. 2008. ‘V. Cook’s multi-competence and its
consequences for SLA research and L2 pedagogy’, Research Reports of
the Department of International Studies (The Department of International
Studies, Faculty of Humanities and Economics, Kochi University) 9,
109–128.

Murahata, G. and Murahata, Y., forthcoming. Dainigengoyuza no Kotoba to

Kokoro [L2 users’ language and mind].
Murahata, Y. 2006. ‘What do we learn from NNEST-related issues? Some
implications for TEFL in Japan’, The Language Teacher 30 (6), 3–7.
Murahata, Y. 2007. ‘Effects of English learning on perception in Japanese
learners of English: a pilot study’, Bulletin of Shikoku Society of English
Language Education 27, 53–64.
Murahata, Y. 2012. A Study of the Effects of English Learning on Cognition in
Japanese Learners of English from a Multi-competence Perspective. PhD disserta-
tion, Newcastle University, UK.
Murahata, Y. and Murahata, G. 2008. ‘Eigo-no gakushu-ha nihongo-no
goijisho-wo kaeru: maruchikonpitensu kenkyu-ga shisasuru mono
[Effects of L2 English learning on L1 Japanese lexicon: what multi-
competence studies tell us]’, Eigo Kyoiku: The English Teachers’ Magazine
56 (11), 14–17.
Murphy, V. A. and Pine, K. J. 2003. ‘L2 influence on L1 linguistic representa-
tions’, in V. J. Cook (ed.), pp. 142–167.
Nicoladis, E., Pika, S. and Marentette, P. 2009. ‘Do French–English bilingual
children gesture more than monolingual children?’, Journal of
Psycholinguistic Research 38, 573–585.
Nisbett, R. 2003. The Geography of Thought: How Asians and Westerners Think
Differently . . . and Why. New York: Free Press.
Odlin, T. 2003. ‘Cross-linguistic influence’, in C. Doughty and M. Long (eds),
Handbook of Second Language Acquisition, pp. 436–486. Oxford: Blackwell.
Odlin, T. 2005. ‘Crosslinguistic influence and conceptual transfer: what are
the concepts?’ Annual Review of Applied Linguistics 25, 3–25.
Ortega, L. 2005. ‘For what and for whom is our research? The ethical as
transformative lens in instructed SLA’, Modern Language Journal 89,
427–443.
Education.
Ortega, L. 2011. ‘SLA after the social turn: where cognitivism and its
alternatives stand’, in D. Atkinson (ed.), Alternative Approaches to Second
Language Acquisition, pp. 167–180. London: Routledge.
Ortega, L. 2012. ‘How relevant is SLA research for language teaching?
Choosing between application and relevance’, in B. Hinger,
E. M. Unterrainer and D. S. Newby (eds), Sprachen Lernen?, pp. 24–38.
Vienna: Praesens Verlag.
Ortega, L. 2013. ‘SLA for the 21st century: disciplinary progress, transdisci-
plinary relevance, and the bi/multilingual turn’, Language Learning
63 (S1), 1–24.
Ortega, L. and Carson, J. G. 2010. ‘Multi-competence, social context, and L2
writing research praxis’, in T. Silva and P. K. Matsuda (eds), Practicing
Theory in Second Language Writing, pp. 48–71. West Lafayette, IN: Parlor
Press.

Osterhout, L., Poliakov, A., Inoue, K., McLaughlin, J., Valentine, G.,
Pitkanen, I., Frenck-Mestre, C. and Hirschensohn, J. 2008. ‘Second-
language learning and changes in the brain’, Journal of Neurolinguistics
21, 509–521.
Pavlenko, A. 2000. ‘L2 influence on L1 in late bilingualism’, Issues in Applied
Pavlenko, A. 2003. ‘ “I feel clumsy speaking Russian”: L2 influence on
L1 in narratives of Russian L2 users of English’, in V. J. Cook (ed.),
pp. 32–61.
Pavlenko, A. 2005. Emotions and Multilingualism. Cambridge University
Press.
Pavlenko, A. (ed.) 2006. Bilingual Minds: Emotional Experience, Expression and
Representation. Clevedon, UK: Multilingual Matters.
Pavlenko, A. (ed.) 2011a. Thinking and Speaking in Two Languages. Clevedon,
Pavlenko, A. 2011b. ‘Introduction: bilingualism and thought in the 20th
century’, in A. Pavlenko (ed.), pp. 1–28.
Pavlenko, A. 2011c. ‘(Re-) naming the world: word-to-referent mapping in
second language speakers’, in A. Pavlenko (ed.), pp. 198–236.
Pavlenko, A. 2011d. ‘Overview of the field’, in A. Pavlenko (ed.),
pp. 237–257.
Pavlenko, A. 2014. The Bilingual Mind. Cambridge University Press.
23, 190–214.
Pavlenko, A. and Malt, B. 2011. ‘Kitchen Russian: cross-linguistic differ-
ences and first-language object naming by Russian–English bilinguals’,
Petitto, L. A., Berens, M. S., Kovelman, I., Dubins, M. H., Jasinska, K. and
Shalinsky, M. 2011. ‘The “Perceptual Wedge Hypothesis” as the basis for
bilingual babies’ phonetic processing advantage: new insights from
fNIRS brain imaging’, Brain and Language 121 (2), 130–143.
Piller, I. 2002. Bilingual Couples Talk: The Discursive Construction of Hybridity.
Porte, G. 2003. ‘English from a distance: code-mixing and blending in the
L1 output of long-term resident overseas EFL teachers’, in V. J. Cook (ed.),
pp. 103–119.
Queen, R. M. 2001. ‘Bilingual intonation patterns: evidence of language
change from Turkish–German bilingual children’, Language in Society 30,
55–80.
Saalbach, H. and Imai, M. 2012. ‘The relation between linguistic categories
and cognition: the case of numeral classifiers’, Language and Cognitive
Processes 27, 381–428.

Sasaki, M. 2005. ‘The effect of L1 reading processes on L2: a crosslinguistic

comparison of Italian and Japanese users of English’, in V. J. Cook and
B. Bassetti (eds), Second Language Writing Systems, pp. 289–308. Clevedon,
Satterfield, T. 1999. Bilingual Selection of Syntactic Knowledge: Extending the
Principles and Parameters Approach. Dordrecht: Kluwer.
Satterfield, T. 2003. ‘Economy of interpretation: patterns of pronoun
selection in transitional bilinguals’, in V. J. Cook (ed.), pp. 214–233.
Schmid, M. S. 2013. ‘First language attrition’, Linguistic Approaches to
Scott, V. M. 2010. Double Talk: Deconstructing Monolingualism in Classroom
Second Language Learning. Upper Saddle River, NJ: Prentice Hall.
Segall, M. H., Campbell, D. T. and Herskovits, M. J. 1963. ‘Cultural
differences in the perception of geometric illusions’, Science 139,
769–771.
Spelke, E. S. and Tsivkin, S. 2001. ‘Language and number: a bilingual
training study’, Cognition 78, 45–88.
Su, I-R. 2010. ‘Transfer of pragmatic competences: a bi-directional perspec-
tive’, Modern Language Journal 94 (1), 87–102.
Su, I-R. 2012. ‘Bi-directional transfer in Chinese learners’ apologizing
behavior’, Concentric Studies in Linguistics 38, 237–266.
Suzuki, E. 2010. L1 to L2 no kankei-ni tuite-no ichi-kousatsu: Chugokujinn
nihonngo gakushuusha no baai [A consideration on the relationship
between L1 and L2: a case of Chinese learners of Japanese]. Poster
session presented at the 10th Annual Conference of Japan Second
Language Association, Gifu, Japan.
Thierry, G., Athanasopoulos, P., Wiggett, A., Dering, B. and Kuipers, J. R.
2009. ‘Unconscious effects of language-specific terminology on pre-
attentive color perception’, Proceedings of the National Academy of Sciences
of the USA 106, 4567–4570.
Timmis, I. 2002. ‘Non-native speaker norms and international English: a
classroom view’, English Language Teaching Journal 56 (3), 240–249.
Tokumaru, Y. 2002. ‘Cross-linguistic influences of L2 English on L1
Japanese in Japanese–English bilinguals’, in Proceedings II Simposio
Internacional Bilinguismo, pp. 399–406. http://ssl.webs.uvigo.es/actas2002/
02/17.%20Yuki%20Tokumaru.pdf
Tokumaru, Y. 2005. An Empirical and Experimental Study on the Cross-linguistic
Relationships between Japanese ‘Katakana-Go’ and English Words: The Bilingual
Mental Lexicon of Multi-competent Japanese–English Users. Unpublished PhD
dissertation, University of Essex, UK.
Unsworth, S. J., Sears, C. R and Pexman, P. M. 2005. ‘Cultural influences
on categorization processes’, Journal of Cross-cultural Psychology 36,
662–688.

Van Hell, J. and Dijkstra, T. 2002. ‘Foreign language knowledge can influ-
ence native language performance in exclusively native contexts’,
Psychonomic Bulletin and Review 9, 780–789.
Ventureyra, V. A., Pallier, C. and Yoo, H.-Y. 2004. ‘The loss of first language
phonetic perception in adopted Koreans’, Journal of Neurolinguistics 17 (1),
79–91.
Videsott, V., Herrnberger, B., Hoenig, K., Schilly, E., Grothe, J., Wiater, W.
and Spitzer, M. 2010. ‘Speaking in multiple languages: neural correlates
of language proficiency in multilingual word production’, Brain and
Language 113, 103–112.
Mouton.
Wierzbicka, A. 2014. Imprisoned in English. Oxford University Press.
Zareva, A. 2010. ‘Multi-competence and L2 users’ associative links: being
unlike nativelike’, International Journal of Applied Linguistics 20, 2–22.

Index
Accessibility Hierarchy, 54, 218 coordinate, 9, 253, 254

accommodation, 479, 492–494 numbers of, 19
acculturation, 471, 536 Birdsong, David, 12, 60, 63
age of onset, 13, 26, 69, 212, 341, 365, 366, 367 Bloomfield, Leonard, 5, 65
aphasia, 506, 507 Boroditsky, Lea, 32, 262, 263, 266, 366, 380
apologising, 300–301, 311–316 brain, 34, 98, 104, 137, 212, 339, 342, 343, 344,
Arabic, 147, 148, 426, 468, 510 346, 381, 383, 393, 394, 432, 511, 522, 523,
Aronin, Larissa, 144, 146, 153 529, 535
aspect hypothesis, 392 Brown, Amanda, 30, 37, 56, 57, 60, 284, 286,
Athanasopoulos, Panos, 31, 32, 34, 51, 251, 264, 391–392
268, 356, 362, 364, 366, 523, 527 Brown, Roger, 379
attrition, 10, 57, 129, 338, 348, 509 Brutt-Griffler, Janina, 3
and migration, 338 Bybee, Joan, 186
and neurophysiology, 343–346 Byram, Mike, 175, 451, 452, 457, 458
neurocognitive approaches, 346–348
Ausbau, 106 Canagarajah, Suresh, 6, 154, 325, 478, 481
Australia, 322, 472 Cantonese, 366, 470
Catalan, 265
Barsalou, Lawrence, 252, 393 categorisation, 32, 358, 362, 379, 383, 386
Basque Center on Cognition, Brain and Centre for Research on Bilingualism, 81
Language, 80 Cherokee, 324, 325, 329
Bassetti, Benedetta, 32, 248, 264, 467 Chinese, 299, 300, 301–303, 470, 471, 508, 527
Berlin, Brent, 370, 379 Chomsky, Noam, 2, 16, 18, 39, 77, 78, 165, 183,
Bernstein, Basil, 13, 171 188, 209, 229, 358, 436, 533, 539
Bialystok, Ellen, 8, 16, 36, 42, 112, 174, 356, 359, citation analysis of multi-competence, 82–83
428, 432 Clahsen, Harald, 232, 236, 347
bilingual cognition, 31–33, 356, 513 Clausal Typing Hypothesis, 210
Bilingual Creative Education (BCE), 434 Clyne, Michael, 19, 346
bilingual immersion, 368 code-switching, 87, 135, 175, 211, 366, 424, 510,
bilingual mode, 118 513, 535
bilingual perspective, 1, 51, 228, 270 cognition, 16, 39, 130, 248, 264, 393–395
bilingualism, 5, 79, 80, 83, 88, 91, 252 cognitive development, 207–209
and IQ, 172, 357 cognitive flexibility, 429, 432
balanced, 66 cognitive routines, 344
cognitive advantages, 34, 89, 111, colour, 31, 269, 270, 358, 359, 360, 362, 363, 364,
174, 413 379, 382, 383, 386, 535
consecutive, 423 Common European Framework of Reference for
definition, 64–67, 249, 254, 406 Languages, 176, 481
disadvantages of, 357 communicative competence, 166, 169, 188, 195,
early, 86, 432 318
in psychology, 88–90 comparisons, 40–41, 61–64, 84, 232, 291, 340,
sequential, 103, 423 420, 453, 525
simultaneous, 103 Complexity Theory, 125
studies, 83–87 compliments, 299
subgroups, 85 concept-mediation model, 256
Bilingualism and Emotions Questionnaire concepts, 194, 248–270, 368, 369, 378, 394
(BEQ), 466 conceptual memory, 427
bilinguals, 79 conceptual restructuring, 177
bimodal, 360 conceptualisation, 392
compound, 253 Contrastive Analysis, 132

560 Index
convergent thinking, 420, 426, 428 frequency, 289, 291

Cook, Vivian, 2, 8, 9, 16, 29, 36, 51, 54, 60, 77, 80, manner-fog, 283
81, 90, 99, 125, 129, 142, 153, 169, 183, semantics of, 290
207, 250, 299, 330, 338, 355, 361, 403, 405, Greek, 362, 383
445, 446, 451, 461, 479, 504 Green, David, 112, 356, 431
creativity, 412, 540 Grosjean, François, 12, 50, 60, 79, 118, 249, 355,
Critical Period Hypothesis, 212, 227, 232 482, 490, 504, 506, 513, 522
Cross-Cultural Speech Act Realization Project, 303 Gujarati, 506
crosslinguistic influence. See transfer Gullberg, Marianne, 56, 60, 65, 276, 280, 284, 288
Cumulative Enhancement Model for Language
Acquisition (CEM), 215 Hall, Joan Kelly, 99, 109, 184, 195, 529
Halliday, Michael, 185, 199
De Bot, Kees, 50, 128, 134, 341, 484, 510, 529 Hawaiian, 329, 333
De Groot, Annette, 79, 254, 266, 530 Hindi, 288
de Saussure, Ferdinand, 17, 153 Hirata-Edds, Tracy, 329, 331
Della Rosa, Pasquale, 34, 112 Hopper, Paul, 132, 135
Dewaele, Jean-Marc, 34, 406, 430, 468, 513 Hungarian, 215
directionality, 219, 278 Hymes, Dell, 165–166, 188–190, 199, 299, 480
discourse completion tests, 302, 310
divergent thinking, 412, 420, 423, 426 implicit knowledge, 105
Dominant Language Constellation (DLC), 10, incremental processing, 230
154, 503 integration continuum, 8, 361
and multi-competence, 160 intercultural awareness, 455
Dörnyei, Zoltan, 453 intercultural competence, 176, 452
Dutch, 58, 137, 238, 239, 255, 284, 288, 464 interlanguage, 2, 13, 292
Dynamic Systems Theory (DST), 10, 92, 115, 131, Invented Alien Creature test, 423
342 Ireland, 20, 328
Irish, 328
effects of bilingualism on cognition, 361–363 Israel, 298, 300, 409
effects of culture, 463–464 Italian, 262
electroencephalography (EEG), 212, 227
embodied cognition, 376, 393–395 Japanese, 29, 34, 214, 268, 278, 283, 362, 391,
emergentism, 131, 132, 133, 134, 241, 529 469, 513
emotion, 426, 461–479 Jarvis, Scott, 10, 55, 285, 358
emotion words, 466 Jenkins, Jenny, 479
Emotional Intelligence, 407
Emotional Patterns Questionnaire, 472 Karuk, 328
endangered languages, 325, 326, 329, 332–333 Kasai, Chise, 251, 363, 365
English, 4, 6, 11, 14, 15, 20, 28, 54, 68, 86, 186, Kazakh, 215
187, 189, 208, 214, 233, 261, 268, 281, 299, Kellerman, Eric, 133, 280, 495
362, 380, 386, 391, 465, 486, 491 Kendon, Adam, 276, 280
English Lingua Franca, 478–497 Kharkhurin, Anatoliy, 174, 420, 423, 425, 429, 434
Ervin-Tripp, Susan, 254, 265, 357 Korean, 189, 210
ethnography of communication, 188 Kramsch, Claire, 101, 166, 167, 455, 457
event-related brain potentials (ERP), 347, 382, 527
Ewert, Anna, 33, 251, 387 L2 user. See second language user
executive control, 356, 358, 360, 364, 365, 367, Labov, William, 13, 67, 169, 252
368, 369 Ladin, 111, 112
eye-tracking, 29, 42, 227, 237, 347, 377, 382, 393, Lambert, Wallace, 8, 16, 66, 69, 78, 79, 254, 357,
396, 527 423, 432
language
Faculty of Language, 2, 207 meaning of, 17, 18, 195, 196, 248, 405, 529,
Farsi, 425, 429 533, 535
Fishman, Joshua, 322, 329 meanings of, 129, 153
flanker task, 111, 360, 365 language awareness, 105, 451–453
Flynn, Suzanne, 207, 211, 213 Language Awareness Forum, 453
fMRI, 103, 112, 212, 384 language biographies, 102
Foreign Language Anxiety (FLA), 405, 410 language identity, 323–325
Foreign Language Enjoyment, 411 language reversion, 345
Franceschini, Rita, 80, 99, 109, 114, 195 language revitalization, 322, 323, 331–332
French, 253, 282, 287, 411, 454 language shift, 321–322
language teaching, 37, 165, 289, 430–433,
Garden Path Model, 229, 230, 233 445–459
gender, 32–34, 262, 264, 266, 267 and native speaker, 14, 35
German, 149, 216, 219, 239, 262, 282, 487, 507 and native speaker teacher, 14, 330
Germany, 392 testing, 293
gestures, 276–293, 299 language-mediated concept activation (LMCA),
and speech, 277–278 427
cross-cultural, 278–280 Latin, 17, 68, 98, 106, 147, 148, 405, 513

Index 561
Latvian, 507 objects, 6, 32, 251, 262–263, 265, 358, 362

Levelt, Willem, 17, 129, 380, 389 Obler, Lorraine, 136
Levinson, Stephen, 31, 185, 300, 379, 380, 539 Odlin, Terence, 52, 133, 284
lexical concepts, 248–270 onset of dementia, 361
lexicon, 17, 127, 154, 210, 465–466 Ortega, Lourdes, 4, 12, 62, 115
lexis, 485–488
Li Wei, 168, 175, 193, 196, 510, 539 Paradis, Michel, 134, 136, 506
linguistic relativity, 16, 17, 31, 41, 176, 250, parameter setting, 210, 220
260–265, 356, 358, 361, 369, 376, parsing. See syntactic processing
377, 396 path, 10, 33, 279, 282, 290, 299, 386, 387, 391
Lucy, John, 31, 32, 261, 358 gestures, 282
Pavlenko, Aneta, 17, 31, 78, 91, 176, 263, 378,
Macaro, Ernesto, 449, 451 424, 461, 473
Malt, Barbara, 265, 380, 381 personality, 403–415, 429
Maltese, 151 Peter, Lizette, 324, 329, 332
Mandarin, 281, 285, 467 Polish, 386, 387, 391, 395, 408, 461, 466,
Māori, 332 474, 514
Mauranen, Anna, 479, 485, 489 Portuguese, 474
McNeill, David, 277, 278, 283, 285 pragmatic competence, 299, 319
Minimalist Program, 210 proficiency, 40, 135, 212, 363, 364, 406, 412, 414,
minority languages, 173, 321 422, 431, 538
Mohawk, 324
monolingual, 40, 86, 92, 116, 155, 228, 466, 521, Quick Oxford Placement Test, 40, 363
522, 525, 528
monolingual perspective, 1, 38, 89, 227, 228, racial terms, 462
229, 231, 238, 533 Radein Initiative, 355
monolingualism, 78, 89, 195, 347, 533, 537 Regression Hypothesis, 341
dominance, 322 relative clauses, 54, 214, 215, 217, 218, 237, 238
motion, 33, 284, 290, 299, 367, 379, 381, repertoire, 107, 125–131, 146, 192, 194, 197
383–387, 391, 395 requesting, 300, 301–303
endpoints, 390–393 research design, 196, 228, 367, 538
motivation, 430, 458 1 group, 54
multi-competence, 227 2 language, 39, 54
and creativity, 430–433 3 language, 39, 54
and ELF, 494 4 language, 39, 55
and native speakers, 61–63 research methodology, 13, 15, 156, 255, 523, 526
cognitive view, 185–188, 198 research questions, 3, 15, 70
definition, 4, 99, 100, 114, 142, 403 for multi-competence, 27, 28
framework, 286, 292, 293 Revised Hierarchical Model, 257, 258, 528
holistic, 36, 52, 58–61, 70, 80, 126, 137, 338, Roberson, Debbie, 358, 381
403, 446, 502, 534 Roberts, Leah, 235, 238, 241
of the community, 3, 146, 161, 169, 178, 321, Russian, 269, 298, 362, 413, 463, 465, 506
328, 449
Premise 1, 7–11, 37, 38, 56, 58, 130, 286, 326, Saer, David, 172, 357
524 Sapir–Whorf hypothesis. See linguistic relativity
Premise 2, 7–11, 35, 38, 64, 130, 525 Sasaki, Miho, 264
Premise 3, 15–19, 30, 41, 130, 168, 323, 525 satellite-framed languages, 390
multi-competence framework, 355, 370, 445 Schmid, Monika, 10, 340, 461
multi-competence perspective, 449, 452, 538 Scott, Virginia, 448, 456
Multicultural Personality Questionnaire, 408 second language, 6, 415
multiculturalism, 404, 425 and foreign language, 4
multilingualism, 5, 7, 27, 101, 103, 109, 110, 111, second language (L2) learner, 62, 446–448
159, 176, 220, 522 second language (L2) user, 4–7, 62, 339, 446, 450,
and creativity, 179 536
definition, 109 second language research, 1
effects on emotion, 410–414 see SLA research
late, 103 semilingualism, 60, 173,
operationalising, 405–407 sense of self, 454
multilingualism research, 1, 97, 106 Service, Elisabet, 511
multilinguality, 125–131, 143, 144 Shallow Structure Hypothesis (SSH), 232, 347
Murahata, Yoshiko, 30, 32 Simon task, 359, 367
Singapore, 509
Nation vocabulary test, 364 Singleton, David, 144, 502, 512, 514
Native American languages, 328, 329 SLA research, 1, 3, 4, 8, 11, 13, 14, 31, 38, 57, 61,
native speaker, 11, 67, 100, 324, 447, 479, 495 63, 502, 533
monolingual, 10, 14, 38, 227, 525 classical, 15, 27, 39, 58
Navajo, 265, 324 Slabakova, Roumyana, 64
New Linguistic Dispensation, 142 Slobin, Dan, 17, 41, 251, 387–390

562 Index
sociolinguistics, 164, 524 L1 to L2, 105, 314, 391

Spanish, 262, 267, 282, 283, 391, 392 L2 to L1, 28, 29–30, 298, 308
Stern, David, 8 pragmatic, 298
Strong Continuity Hypothesis, 211, 213 translanguaging, 169, 193, 333, 424, 510, 533,
Stroop interference task, 359 539–542
Su, I-Ru, 299, 300
subordinate clauses, 219 UK, 7, 20, 179, 191
successful language learner, 449 Universal Grammar, 3, 5, 61, 207, 208, 210,
swearing, 467, 471, 475 212, 534
Swedish, 33, 288, 392 USA, 279, 283, 325, 472, 528
Swiss German, 106, 506 Ushioda, Ema, 456
syntactic fluency, 207, 217, 220, 221
syntactic processing, 227, 228, 236 Vaid, Jyotsna, 79, 528
models, 229–232 variability, 108
syntax, 206–221, 464, 490 variation, 169–170
verbalization, 268–269
Taiwan, 302 verb-framed languages, 390
Talmy, Leonard, 16, 33, 191, 284, 390 Voice Onset Time (VOT), 28, 340
Theory of Mind (TOM), 250
Thierry, Guillaume, 34, 42, 359, 369, 524 Weinreich, Uriel, 9, 28, 57, 79, 97, 98, 131,
thinking for speaking, 17, 251, 253, 389
284, 387–388, 395 Welsh, 322, 357
Third Quality, 105 Werker, Janet, 515
third space, 455, 458 wh-dependencies, 236
threshold theory, 423 wh-movement, 210, 231
Tokumaru, Yuki, 29 Whorf, Benjamin Lee, 10, 31, 260, 262, 361,
tolerance of ambiguity, 410, 429 377–379
Tomasello, Michael, 18, 186 Wierzbicka, Anna, 17, 462
transfer, 10, 27, 52, 85, 133, 213, 280, 298, 299, working memory, 231, 240, 359
304, 316, 345
conceptual, 390–393 Yee, Nick, 385

Vivian Cook_ Li Wei - The Cambridge Handbook of Linguistic Multi-Competence-Cambridge University Press (2016)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Vivian Cook_ Li Wei - The Cambridge Handbook of Linguistic Multi-Competence-Cambridge University Press (2016)

Uploaded by

Copyright:

Available Formats

The Cambridge Handbook of Linguistic Multi-competence

V I V I A N C O O K is Emeritus Professor of Applied Linguistics at Newcastle University

and Visiting Professor at the University of York. He previously taught Applied

Linguistics, at UCL Institute of Education, University College London. He has

Published online by Cambridge University Press

Genuinely broad in scope, each handbook in this series provides a complete

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

Published online by Cambridge University Press

V I V I A N C O O K is Emeritus Professor of Applied Linguistics at Newcastle University

and Visiting Professor at the University of York. He previously taught Applied

Linguistics, at UCL Institute of Education, University College London. He has

Published online by Cambridge University Press

Genuinely broad in scope, each handbook in this series provides a complete

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

Published online by Cambridge University Press

V I V I A N C O O K is Emeritus Professor of Applied Linguistics at Newcastle University

and Visiting Professor at the University of York. He previously taught Applied

Linguistics, at UCL Institute of Education, University College London. He has

Published online by Cambridge University Press

Genuinely broad in scope, each handbook in this series provides a complete

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

Published online by Cambridge University Press

V I V I A N C O O K is Emeritus Professor of Applied Linguistics at Newcastle University

and Visiting Professor at the University of York. He previously taught Applied

Linguistics, at UCL Institute of Education, University College London. He has

Published online by Cambridge University Press

Genuinely broad in scope, each handbook in this series provides a complete

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

Published online by Cambridge University Press

List of figures page vii

1 Premises of multi-competence Vivian Cook 1

Published online by Cambridge University Press

16 Multi-competence and first language attrition Bregtje Seton

Bibliography of ‘multi-competence’ (and related topics) Goro Murahata,

Published online by Cambridge University Press

1.1 The integration continuum of possible relationships

Published online by Cambridge University Press

3.1 Study designs for the investigation of cross-linguistic

Published online by Cambridge University Press

14.12 Mean number of apology strategies in relation to social

Published online by Cambridge University Press

Larissa Aronin Associate Professor, Oranim Academic College of

Published online by Cambridge University Press

Anatoliy V. Kharkhurin Associate Professor of Psychology, Department

Published online by Cambridge University Press

Published online by Cambridge University Press

This chapter introduces the concept of linguistic multi-competence

1.1 Monolingual and bilingual perspective

https://doi.org/10.1017/CBO9781107425965.001 Published online by Cambridge University Press

1.2 Defining multi-competence

Let us start with the conceptual history of multi-competence. Franceschini

(i) ‘the compound state of a mind with two grammars’

https://doi.org/10.1017/CBO9781107425965.001 Published online by Cambridge University Press

– making clear that multi-competence is not confined to syntax but

(iii) ‘the overall system of a mind or a community that uses more

https://doi.org/10.1017/CBO9781107425965.001 Published online by Cambridge University Press