Serial Verbs Alexandra Y.

Aikhenvald
Visit to download the full and correct content document:
https://ebookmass.com/product/serial-verbs-alexandra-y-aikhenvald/
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

Serial Verbs
OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

O X F O R D S T U D I E S I N T Y P O LO G Y AN D LI N GU I S T I C T H E O R Y
 : Sonia Cristofaro, University of Pavia, William Croft, University of
New Mexico, Nicholas Evans, Australian National University, Heiko Narrog, Tohoku
University, and Eva Schultze-Berndt, University of Manchester


Classifiers: A Typology of Noun Categorization Devices
Alexandra Y. Aikhenvald
Imperatives and Commands
Alexandra Y. Aikhenvald
Serial Verbs
Alexandra Y. Aikhenvald
Auxiliary Verb Constructions
Gregory D. S. Anderson
Pronouns
D. N. S. Bhat
Subordination
Sonia Cristofaro
The Paradigmatic Structure of Person Marking
Michael Cysouw
Adpositions
Claude Hagège
Indefinite Pronouns
Martin Haspelmath
Anaphora
Yan Huang
Reference in Discourse
Andrej A. Kibrik
The Emergence of Distinctive Features
Jeff Mielke
Applicative Constructions
David A. Peterson
Copulas
Regina Pustet
The Noun Phrase
Jan Rijkhoff
Intransitive Predication
Leon Stassen
Predicative Possession
Leon Stassen
Co-Compounds and Natural Coordination
Bernhard Wälchli

     

The World Atlas of Language Structures
edited by Martin Haspelmath, Matthew S. Dryer, David Gil, and Bernard Comrie
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

Serial Verbs

ALEXANDRA Y. AIKHENVALD
Language and Culture Research Centre
James Cook University

1
OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

3
Great Clarendon Street, Oxford,  ,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Alexandra Y. Aikhenvald 
The moral rights of the author have been asserted
First Edition published in 
Impression: 
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
 Madison Avenue, New York, NY , United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 
ISBN ––––
Printed and bound by
CPI Group (UK) Ltd, Croydon,  
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

Preface and acknowledgements

My first personal brush with dauntingly complex serial verb constructions was
through fieldwork on Tariana, an Arawak language in the Vaupés River Basin in
Brazil, and other related Arawak languages. To come to grips with the new facts,
I decided to undertake a typological study of serial verbs.
In  R. M. W. Dixon and I organized an International Workshop on ‘Serial
verb constructions’. The workshop and the ensuing edited volume featured an
introduction by myself, setting out the cross-linguistic parameters of variation of
serial verb constructions, their classification and general issues, with a summary by
Dixon summarizing the parameters of analysis. This was followed by fourteen
contributions featuring indepth analyses of serial verb constructions and similar
structures in languages from Asia, Africa, North, Central, and South America,
Oceania, and New Guinea. The analytic framework—in particular the division of
serial verbs into symmetrical and asymmetrical, and the principles of grammatical-
ization and lexicalization within them—has come to be accepted as standard in the
growing number of hands-on studies of serial verbs across the world. The past few
years have seen new data, new languages, and new systems of serial verb construc-
tions analysed and recognized. Alongside these, there have been revisionist attempts
at diluting the concept of ‘serial verbs’ within the sea of multi-verb combinations of
all sorts. Coming from another angle, there has been distortion of the concept of verb
serialization through misinterpreting sources, misunderstanding arguments, and
taking aim at the essence of typologically informed grammars.
The time is now ripe to systematically take stock of serial verb constructions
world-wide, incorporating new data and new insights. Encouraged by colleagues
and independent reviewers from Oxford University Press, I decided to undertake a
cross-linguistic study of serial verbs, world-wide.
The book can be used both as a sourcebook for further typological investigations,
and a textbook. The discussion is couched within basic linguistic theory (rather than
any of the time-line formalisms), and all generalizations are based on reliable
language facts. The readers and myself share a common purpose—to gain an
understanding of a wide diversity of serial verbs. I have tried to analyse the facts
and present the discussion in the clearest possible way—the complexity of the
systems will speak for itself.
I am indebted to many people, of different continents and backgrounds.
My gratitude goes to all those native speakers who taught me their languages and
their serial verb constructions: José, Jovino, Olívia, Rafael, Leo and the late Cândido,
Graciliano and Ismael Brito (speakers of Tariana of Santa Rosa), Marino, Domingo,
OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

x Preface and acknowledgements

Ismael, Jorge and Batista Muniz (Tariana of Periquitos); Humberto Baltazar and
Pedro Ângelo Tomas (Warekena); Afonso, Albino and João Fontes, Celestino da
Silva, Cecília and Laureano da Silva, and the late Marcília Rodrigues (Baniwa). I am
forever indebted to Pauline Agnes Yuaneng Luma Laki, James Sesu Laki, Jacklyn
Yuamali Benji Ala, Katie Teketay, Jenny Kudapa:kw and Gemaj, and to so many
others who revealed the beauty of their native Manambu (Papua New Guinea).
Tamara Margolina, together with my late mother, Valeria Guerlin, helped me get
some linguistic understanding of double verb constructions in their native Russian.
Invaluable support came from those who gave me comments on earlier versions,
patiently answered my questions on serial verbs and related issues, and provided
me with references and additional sources over the year—Grant Aiton, Ayhan
Aksu-Koç, Felix Ameka, Azeb Amha, Alexander Andrason, Junwei Bai (Abe),
Pier Marco Bertinetto, Olga Blinova, Juliane Böttger, Seino van Breughel, Hilary
Chappell, Luca Ciucci, Andrew Cowell, Eva Csató, Wolf Dietrich, Tony Diller, Gerrit
J. Dimmendaal, Mark Durie, Carola Emkow, Nick Enfield, Diana Forker, Lucía
Golluscio, Valérie Guérin, Antoine Guillaume, Bernd Heine, Bernhard Hurch,
Gwen Hyslop, Emi Ireland, Bart Jacobs, Nerida Jarkey, Cheryl Jensen, Lars Johanson,
Christa König, Adenike S. Lawal, Yongxian Luo, Silvia Luraghi, Lise Menn, Marina
Magalhães, Stephen Matthews, Felicity Meakins, Elena Mihas, Veronica Nercesian,
Rachel Nordlinger, Andy Pawley, Ekaterina Protassova, Nick Reid, Willem de Reuse,
Eduardo Ribeiro, Keren Rice, Malcolm Ross, Aaron Rubin, Hannah Sarvasy, Dineke
Schokkin, Sandy Steever, Kris Stenzel, Anne Storch, Tim Thornes, Pilar Valenzuela,
René van den Berg, Maria Voejkova, Pema Wangdi, John Watters, Nathan White,
Mary Ruth Wise, Kasia Wojtylak, Firew Girma Worku, Defen Yu, and Sihong Zhang.
Invaluable comments on just about every page came from R. M. W. Dixon,
without whose incisive criticism, ideas, and constant support and encouragement
this book would have never come to fruition.
I owe a debt of gratitude to the library staff of James Cook University—and
especially Carolyn Tredrea and Bronwen Forster—whose constant support continues
being crucial for our work. The efficient interlibrary loan system worked like
clockwork—my particular thanks go to Lyn Clark and Bridie Soo. I owe a consider-
able debt to Brigitta Flick and Jolene Overall for carefully reading through the many
drafts of this book and making corrections, and to David Ellis and Jolene Overall for
looking after all things administrative while I was doing the writing. The work was
supported by the Australian Research Council Discovery projects ‘How languages
differ, and why’ and ‘The integration of language and society’, and also my Australian
Laureate Fellowship.
This volume would have never been brought to fruition without encouragement
from Julia Steer and Vicki Sunter, linguistics editors of Oxford University Press.
Their unfailing support and efficiency made it all worthwhile.
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

List of figures and tables

Figures
. Verbs in serial verb constructions: from the more likely to the less likely 
. The order of development of asymmetrical serial verb constructions 
. The emergence of multi-word serial verbs in four Tupí-Guaraní
languages 
. Concurrent grammaticalization scenario for the development of asymmetrical
serial verb constructions 
. The order of historical development of asymmetrical serial verb constructions
by semantic group 
. Verbs in serial verb constructions: from the more likely to the less likely
(repeated from Figure .) 
. The order of development of asymmetrical serial verb constructions
(repeated from Figure .) 
. The emergence of multi-word serial verbs in four Tupí-Guaraní languages
(repeated from Figure .) 
. Concurrent grammaticalization scenario for the development of asymmetrical
serial verb constructions (repeated from Figure .) 

Tables
. Properties of serial verb constructions with non-identical subjects 
. Asymmetrical and symmetrical serial verbs: a comparison 
. Directional and positional markers in Cavineña, and their source verbs 
. Several kinds of serial verbs in one language: contiguity and wordhood 
. Causative constructions in Tetun Dili 
. Asymmetrical and symmetrical serial verbs: a comparison
(repeated from Table .) 
. Serial verbs, clause sequences, and multi-verb predicates with converbs 
OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

Abbreviations and conventions

Abbreviations
 transitive subject
 ablative
 absolutive
 accusative
 additive ‘and’
 allative
 anaphoric
 animate
 anterior
 applicative
 article
 aspect
 assertion
 asymmetrical
 augmented
 causative
 clausal focus
 centrifugal
 classifier
 collective
 comitative
 comitative
 complementizer
 completive
 conditional
 connective
 consecutive
 continuative, continuous
. continuous derivational suffix on coverbs
 contemporary
. content interrogative
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

Abbreviations and conventions xiii

 converb
 dative
 declarative
 deductive evidential
 definite
. definite future
 demonstrative
. demonstrative general
 dependent
 desiderative
 diminutive
 distant mood
 distal
dl dual
 different subject marking
 dual
du dual
 elevational
 emphatic
 epenthetic
 ergative
exc,  exclusive
excl exclusive
 existential
 experiential aspect
f,  feminine
fem,  feminine
 full form
 focus
. focus particle
 frequentative
 future
. future irrealis
 goal
 genitive
 habitual
 immediate
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

xiv Abbreviations and conventions

 impersonal
 imperfective
 imperative
 inchoative
incl inclusive
 indefinite person
. independent pronoun
 indefinite
 indicative
 infinitive
 injunctive
 instrumental
 intentional
 interjection
 irrealis
KBo Keilschrifttexte aus Bodhazköi (Cuneiform texts from Boghazköi)
l l-conjugation in Dyirbal
 linker
 locative
log.addr addressee logophoric
log.sp speaker logophoric
m,  masculine
masc,  masculine
 middle
 minimal
. modal locational
 noun class
 negative
. negative verb
 nonfuture
 nominalizer
 nominative case
 nonfuture
 noun phrase
nsg,  nonsingular
 numeral
.: numeral classifier for animates
 OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

Abbreviations and conventions xv

O object
 object
 oblique
 obligative
p person
 participle
. question particle
 passive
 past (no abbreviation)
 perfective
pers,  person marker
 perfective
 perfective
 past imperfective
pl,  plural
poss,  possessive
 potential
 precontemporary tense
 preposition
 present
. present visual
 probability
 progressive
 prohibitive
. proximal demonstrative
 partitive
 particle
 purposive
. question particle
 realis
.. recent past visual
 reciprocal
 reduplication
 prior reference made
 reflexive
 referential
 relative
OUP CORRECTED PROOF – FINAL, 15/9/2018, SPi

xvi Abbreviations and conventions

. remote past

.. remote past reported
.. remote past visual
S intransitive subject
 sequential
sg singular
sgf singular feminine
sgm singular masculine
sgnf singular nonfeminine
 simultaneous
 sentence patient
 specific
 same subject marking
 subject
 subjunctive
 subordinator
SVC serial verb construction
 symmetrical
TA tense aspect
 telic
../ topical non-subject
. topical object
 transitive
 transitive
 usitative
V verb
V first verb in a serial verb construction
V second verb in a serial verb construction
.. comitative verbal particle
 vertical
 verbal noun

Conventions
Verbs within serial verb constructions are in bold. A translation of a serial verb into English is
accompanied by a literary rendering of the verbs the construction consists of (for instance,
take go). A full term is used instead of an abbreviation (e.g. , , ), for ease
of understanding. Transcription and glossing of examples is that of the original sources.
Language names are presented as the authors of relevant grammars spell them.
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

Serial verbs
The framework

In many languages of the world, a sequence of several verbs act together as one unit.
They form one predicate, and contain no overt marker of coordination, subordin-
ation, or syntactic dependency of any other sort. Such series of verbs are known as
serial verb constructions, or serial verbs for short. Serial verbs describe what can be
conceptualized as a single event. They are often pronounced as if they were one word.
Serial verbs tend to share subjects and objects. They have just one tense, aspect,
mood, and modality value—that is, one part cannot refer to past, and another to present.
The components of serial verbs cannot be negated or questioned separately from the
whole construction. Each component must be able to occur on its own. The individual
verbs within the construction may have the same transitivity values; or the values may be
different. Serial verbs are a powerful means for a detailed portrayal of various facets of
one single happening. They often express grammatical meanings—aspect, direction, or
causation—especially in languages where few other means are available.
Serial verbs are found in many languages, with different typological profiles. They
are prominent in European-based Creole languages, and in isolating languages of West
Africa and of South-East Asia. They have now been recognized in numerous languages
of Oceania and New Guinea (especially those of the Oceanic subgroup of the large
Austronesian family), and of the Americas (including the Amazonian Lowlands). They
have been described for at least a dozen Australian Aboriginal languages, a number of
varieties of colloquial Arabic, Syriac Aramaic, Dravidian languages of India, numerous
Tibeto-Burman languages, a few languages of north-east Europe, and a number of
extinct Indo-European languages (including Hittite and Classical Armenian).

. What serial verbs are like

Serial verbs have a plethora of meanings and functions. Let’s start with some
examples. In ., from Tetun Dili, an Austronesian language from East Timor, the
serial verb expresses the direction of motion: ‘throw come’ means ‘throw over here’.
The serial verb construction has directional meaning. Here and elsewhere compo-
nents of a serial verb construction are in bold.

Serial Verbs. First edition. Alexandra Y. Aikhenvald.

© Alexandra Y. Aikhenvald . First published  by Oxford University Press.
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

. tuda bola mai Tetun Dili

throw ball come
‘Throw the ball over here’ (lit. throw come) (Hajek : )
The serial verb in ., from Ijo, an Ijoid language from Nigeria, contains the verb
‘take’: this verb allows the instrument ‘knife’ to be used with the verb ‘cut’. This is an
instrumental serial verb construction.
. Erí ogidi akí-ní indi pei-mí Ijo
he machete take-/ fish cut-/
‘He cut a fish with a machete’ (lit. take cut) (McWhorter : )
One verb can refer to the effect of the other, as in ., from Taba, an Austronesian
language from Indonesia. The idea of another animal biting the pig dead is expressed
with the series of verbs ‘bite die’. The marker of realis occurs once in the verbal
complex—characterizing it as a whole—and follows the last verb. This is a cause-
effect serial verb construction.
. n=babas welik n=mot do Taba
sg=bite pig sg=die 
‘It bit the pig dead’ (literally: bite die) (Bowden : )
A serial verb construction may refer to a sequence of interconnected actions. In ., from
Alamblak, a Sepik Hill language from Papua New Guinea (Bruce : ), ‘getting’ and
‘putting’ are subsequent to each other. This is a sequential serial verb construction.
. wa-rim-ak-hɨta-n-m Alamblak
--get-put-sg-pl
‘Get them on a level plane toward me (and) put them up there’ (lit. get put)
(Bruce : )
Verbs within a serial verb construction can form an idiomatic collocation. In many
West African languages the concept ‘believe’ is expressed with a serial verb whose
overall meaning cannot be easily inferred from the meanings of its parts, as in Ewe xɔ se
‘believe’, literally, ‘get hear’, or Akan gye di ‘believe’, literally, ‘get eat’ (Ameka : ).
A serial verb construction can consist of more than just two verbs. The sentence in
., from Tariana, an Arawak language from Northwest Amazonia, contains a serial
verb with three components—‘take’, ‘make cross’, and ‘make stand’. The meaning of
the serial verb is ‘bring across’.
. phia-ka phita pi-thaketa pi-eme Tariana
you-.. sg+take sg-cross+ sg-stand+
ha-ne-na hyapa-na-nuku
--: hill-:-../
‘It was you who brought that mountain across (lit. take make.cross make.stand)
(the river) to the other side’ (said the king)
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Defining a serial verb 

A serial verb in Sranan, a Dutch-based Creole, can be even longer. The sequence of
verbs ‘run’, ‘go’, ‘take’, ‘carry’, go’, and ‘give’ in . forms one serial verb construction.
. Lon go teki a buku tyari go gi a leriman Sranan
run go take the book carry go give the teacher
‘Run (and) fetch the book (and) take it to the teacher’ (run go take carry go give)
(Sebba : )
A serial verb is often best translated with a monoverbal clause into a language which
lacks serial verbs, such as English or Portuguese. During a fieldwork session with
Tariana speakers, I had difficulties in taking in the fact that, to render the Portuguese
verb trazer ‘bring’ into Tariana, one has to use a serial verb -hita -inu ‘get come’. My
teacher, the late Graciliano Sanchez Brito, came to my aid, explaining: ‘It is not like
Portuguese, we just cannot say it with one verb’.
Rendering some types of serial verbs into English can be a challenge. In her
analysis of sequential serial verbs in Gurr-goni, Green (: ) remarks how it
is ‘almost impossible to give a monoclausal translation for these Gurr-goni
examples—English requires one of the verbs to be subordinated to or co-ordinate
with the other’. The serial verb constructions in Gurr-goni have the intonational
contour of a single clause, they share tense, mood and polarity value, and ‘however
hard it may be to express in English, the events being described are thought of as one
event, rather than two separate actions or events’. In a similar vein, in his discussion
of serial verb constructions in Lao, Enfield (a: ) remarks that ‘it is impossible
to reflect in the English translation the fact that the . . . elements [of a serial verb
construction] are each unmarked verbs of similar status’. A translation will inevitably
contain the word and. But this does not make serial verbs into coordinate clauses.
Serial verb constructions—diverse as they are—always consist of several verbs
which form one predicate. Verb serialization can be viewed as a grammatical
technique deployed to express a plethora of meanings. How to define a serial verb
is our next question.

. Defining a serial verb

A serial verb construction will have the following properties.
A. A serial verb construction (SVC) consists of two or more verbs, each of
which could also function as the sole verb in a clause.
A verb may have a different meaning when used alone and when used in an SVC,
but the meanings should be relatable. The verb has to be able to stand alone.
B. There is no mark of dependency—such as coordination, subordination, or
dependency of any sort—between the verbs within a serial verb construction.
It is important to carefully distinguish between serial verb constructions and
other constructions consisting of more than one verb which may have similar
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

(but never identical) features. There is typically a complex of grammatical and

discourse properties which distinguish serial verb constructions from consecu-
tive, subordinate, purposive, and other kinds of verb sequences.
C. A serial verb construction is monoclausal—it functions as a single predicate.
C. Grammatical categories which apply to a monoclausal predicate in a
given language will have the complete serial verb construction in their
scope. These may include tense, aspect, evidentiality, modality, mood,
reality status, illocutionary force, manner adverbs, and also markers of
subordination (including those of relative clauses).
C. Typically, only a complete serial verb construction—and not one of its
components—can be negated or questioned.
C. A serial verb construction will be repeated as a whole. When used in a
response to a question, the reply will not consist of just one component
verb: it will contain the full construction.
C. A serial verb construction cannot be felicitously split into a number of
coordinated clauses.
C. In general, a serial verb construction will fall within one intonation
contour and is pronounced as one verb would be, so that no pause is
likely in the middle of it.
D. The serial verb construction itself will have its own transitivity value.
Verbs which make up a serial verb construction may vary in their transitiv-
ities. An SVC has an overall transitivity which will depend on the transitivity
of the component verbs and the type of serial verb construction.
E. There is usually at least one core argument shared by all the verbs in a serial
verb construction.
The shared argument is typically the syntactic subject (covering the transitive
subject labelled A and the intransitive subject labelled S). Objects and
obliques can also be shared.
F. The serial verb construction is conceived as describing a single event.
The event described by a serial verb construction may be simple or complex.
It may cover a set of interconnected activities, or states.

A prototypical serial verb construction in a given language will have all the properties,
from A–F. Features A–C help distinguish serial verbs from other sequences of verbs, or
verb-like elements. To show that a given sequence of verbs is indeed a serial verb
construction, one needs to prove that the sequence forms one predicate and is mono-
clausal. Strict tests will be applied to show that the serial verb has one value for all the
predicate categories. A single caveat is in order. A contiguous serial verb construction
will be pronounced in the same way as a monoverbal predicate. But a discontinuous
serial verb may not be, and the parameter C may not universally apply.
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Defining a serial verb 

A prototypical serial verb will share all the core arguments—the subjects and the
objects—and also obliques, and constitute a cohesive and tightly-knit representation
of one event. And indeed, every language with serial verbs will have SVCs with the
same subject. But just occasionally, a language will have a subtype of serial verbs
whose components describe different facets of the same event, and yet no arguments
are shared. Then, parameter D will only partially apply.
The semantic unity of a serial verb depicting one event is a further matter. Showing
that a serial verb refers to one single event may be unproblematic in some instances—
as in . from Tetun Dili and . from Ijo—but more difficult in others. Sequential
serial verbs—like the one in . from Alamblak—may be analysed as a closely-knit
sequence of subevents, or as one ‘macroevent’. But how to make this foolproof? Serial
verbs have to be defined in the first place based on their formal features; a purely
semantic definition of serial verbs—if limited to just parameter F—may run into
problems to do with the hard-to-pinpoint notions.
This suggests a scalar, or continuum-type approach, to serial verbs—which can
conform to a prototype (A–E) to greater or lesser extent.
The components of serial verb constructions may always have to occur next to
each other—as in . and .. Or other constituents may intervene between them—as
in .– and .. Some verbal categories may have to be marked on every component
in a serial verb, as is the tense/aspect in . and person in ., ., and .. Or a
category can be marked once per construction—an example is the realis in .. All
components of a serial construction may share their subject, as in .– and .–. Or
they may share another argument: in . the object of the first component (‘bite’) is
the same as the subject of the second one (‘die’).
Serial verb constructions may not be limited to just some forms and contexts.
A concatenation of verbs in American English come play! is not a serial verb because
this is limited just to imperative and a few other constructions—one can say I want
you to come play but not *he came played.1
In their functions and meanings, serial verb constructions may appear similar to
multi-verb constructions which contain subordinating or coordinating linkers, or
dependent verb forms such as converbs, or complex verb forms consisting of
auxiliaries plus dependent verb forms. As Matisoff (: ) put it, serial verbs
‘serve to provide in a uniform way the sort of information that in the surface
grammar of languages like English is handled by a formally disparate array of
subordinating devices: complementary infinitives, -ing complements, modal auxil-
iaries, adverbs, prepositional phrases, even whole subordinate clauses’. Throughout
this book, we will point out differences and commonalities between serial verbs and
other verb sequences, and their interactions in the history of languages.
Serial verb constructions come in a variety of guises. They may consist of several
phonological and grammatical words, as in .–, and .–. Or they may form one
word, as in .. This can be alternatively referred to as ‘root serialization’.2 Single-word
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

serial verb constructions have to be distinguished from verbal compounds—limited,

lexicalized, and unproductive combinations whose meaning is often unpredictable
and which can be listed exhaustively. Korean has a limited number of semantically
unpredictable verb–verb combinations, such as mi-tayta ‘lay, throw, shift’ (made up
of roots mil ‘push’ and tay ‘attach’) or pwuth-tulta ‘take hold of ’ (made up of roots
pwuth ‘stick’ and tul ‘hold’). These can be listed exhaustively and are idiomatic (Sohn
: ). Such compounds are not serial verbs.
We distinguish two broad types of serial verb constructions, in terms of their
composition.
• A  serial verb construction involves two or more verbs of
different status. The ‘major’ component can be virtually any verb; it is chosen
from a semantically and grammatical unrestricted class, and can be considered
the ‘head’ of the construction. The ‘minor’ component is chosen from a limited
and closed subclass of verbs of a certain semantic set. Among semantic values a
minor verb may specify are:
(i) direction—coming or going, ascending or descending, moving across,
etc.—or posture and stance such as sitting or standing;
(ii) aspect, extent, or change of state, covering progressive, continuative, com-
pletive, habitual, and so on;
(iii) modal values of obligation, necessity, probability, and possibility;
(iv) phasal meanings such as starting and finishing;
(v) adding an argument and increasing valency.
An example of an asymmetrical serial verb with directional meaning is at .. The
asymmetrical serial verb in . has instrumental meaning, with the minor verb ‘take’.
• A  serial verb construction combines verbs of any semantic type.
None of the components can be considered the ‘head’. Examples . and . are
typical instances. The only restriction on combining the components into one
construction may have to do with semantic plausibility of the whole. For
instance, in White Hmong a symmetrical serial verb construction dhia tshov
(dance blow) is acceptable, because the actions of dancing and blowing bamboo
pipes are seen as closely knit together in the Hmong musical repertoire. But
‘dance’ and ‘listen to music’ are normally viewed as distinct events, and the two
verbs cannot form one serial verb construction (Jarkey : –).
Every language with serial verb constructions has asymmetrical serial verbs. No
languages with symmetrical serial verbs and without asymmetrical ones have been
identified thus far.
Further parameters of variation in serial verb constructions are:
I. C versus - of components. Verbs which form a serial
verb construction may have to be next to each other, or another constituent may be
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Serial verbs in linguistic history 

allowed to intervene between them. The serial verb in . is strictly contiguous; the
verbs in .– are not.
II. W  : components of a serial verb construction may
or may not form independent grammatical or phonological words. The serial verb in
. forms one grammatical word; serial verbs in .– and .– consist of several.
III. M of grammatical categories in a serial verb construction: verbal
categories—such as, for instance, person of the subject and object(s); tense, aspect,
modality; negation; or valency changing—may be marked just once per construction
(‘ ’); or can be marked on every component (‘ ’).
Tense/aspect in . and person in . and . receive concordant marking within the
serial verb. Realis in . is marked just once per serial verb construction.
Verb serialization may be fully productive. Or it may be limited to just a few
semantic types. A language may have several kinds of serial verbs—varying in their
contiguity, wordhood, and meanings. Some verbs—such as those of motion and
direction—are more likely to occur in serial constructions than others. In many
languages, a serial verb cannot consist of verbs denoting states. In the course of
language history, the minor verb in an asymmetrical serial verb construction may
undergo grammaticalization and become an affix, an auxiliary, or a preposition.
A symmetrical serial verb construction will tend to lexicalize, and develop into one
unit, no longer segmentable—losing its status as a serial verb. The many types and
guises of serial verbs across languages, their formal, semantic, and pragmatic features,
and their histories is what this book is about. We now turn to a brief history of
linguists coming to grips with serial verb constructions.

. Serial verb constructions in the history of linguistics

The concept of verb serialization is a fairly recent arrival on the linguistic scene. Up
until the late nineteenth century, only the linguistic categories prominent in classical
Indo-European languages were, by and large, accorded due status and investigated in
some depth. Serial verb constructions were not among these. It has taken linguists
some time to acknowledge the existence of serial verb constructions and to find them
a place in an analytic framework.
In the second half of the nineteenth century linguists noted that, in a number of
African languages, several verbs combine in a way distinct from what we see in familiar
European languages. Adding a verb to another verb has a similar effect to an adverb of
time and manner, or a preposition of instrument, motion, or position, in an Indo-
European, or a Semitic language. As J. G. Christaller (: ), a German missionary,
put it in his renowned grammar of Asante and Fante, from the Kwa language family,
‘many verbal notions that are expressed with a simple verb in English and other
European and Asiatic languages, are expressed by syntactical combinations of verbs’.3
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

The phenomenon of more than one verb in a row without any mark of syntactic
link was recognized in many a classic work. In his grammar of Ewe, another Kwa
language, Westermann (: ) describes serial verbs in the language:
a peculiarity of the Ewe language is that we often find a row of verbs one after the other. The chief
features of this are that all verbs stand next to each other without being connected, that all have
the same tense or mood . . . In English these consecutive verbs are partly rendered by composite
sentences. But very often several Ewe verbs may be expressed by a single verb in English.
The explanation for this is that the Ewe people describe every detail of an action or
happening from beginning to end, and each detail has to be expressed by a special verb: they
dissect every happening and present it in its several parts, whereas in English we seize on the
leading event and express it by a verb, while subordinate events are either not considered or
rendered by means of a preposition, adverb, conjunction, or prefix of the verb.

Otto Dempwolff, the author of the classic grammar of Jabêm, an Oceanic language
spoken in Papua New Guinea (), recognized serial verb constructions—which he
termed Reihensatz (serial sentence)—and described them as follows:
The representation of several events, usually only two, can be combined into a new represen-
tation, similar to the way new combinations are represented in German by means of verbal
prefixes, as in weichen ‘to move’ vs ausweichen ‘to move out [of the way]’ . . . To this end, full
verb forms are placed in sequence.4

It was not until  that the term ‘serial verb’ was coined, by Balmer and Grant in
their grammar of Fante Akan (: –). In their own words,
there is . . . one usage which is a distinctive feature of Fante verbs, viz. the use of double or
twofold verbs, as, gye . . . dzi, to believe. This is due partly (a) to the tendency of the language to
use vivid figurative expressions and partly (b) to the habit of analysing an action into its
component parts . . . These verbs may be termed serial verbs. (pp. , )

The term ‘serial verbal construction’ was reintroduced by John Stewart (), in the
paper ‘Some restrictions on objects in Twi’.5 From then onwards the terms ‘serial
verb construction’ and ‘serial verb’ (used by Stahlke ) have gradually won
general acceptance. A few alternative terms appear in the literature—including
‘verb concatenations’ (Matisoff ,  work on Lahu, a Tibeto-Burman lan-
guage), ‘tandem patterns of verb expressions’ (Senft , in his grammar of Kilivila,
an Oceanic language), ‘strings of verbs’ (Williamson  on Ijo; and Mitchell 
on Colloquial Egyptian Arabic), ‘verbal chains’ (Voorhoeve , on Sranan), and
‘verb series’ (Welmers ). Some serial verbs were described without using the
term—in his grammar of Dyirbal, an Australian language, Dixon (: –) first
referred to them as verb complexes, labelling them ‘serial verbs’ in his subsequent
work (a, , ).
The studies of West African languages underwent a veritable boom in the s
and early s, with serial verb constructions moving into the general spotlight.
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Serial verbs in linguistic history 

Serial verbs were also recognized in languages of South-East Asia, and then for
Creoles and also for Oceanic languages.6 Once this theoretical window was open,
many linguists attempted to fit serial verbs into the familiar framework of two clauses
coordinated together.
The then current linguistic transformational-generative paradigm stated operated
on one-verb-one-clause principle. Serial verbs—sequences of more than one verb—
had to be derived from underlyingly coordinated sequences of clauses, with a
subsequent deletion of co-referential subjects and the coordinator. This is how
Stewart (), and a fair few others (including Bamgbos̩e  and George ),
analysed serial verb constructions.
The assumption ‘that simple sentence may have only one finite verb, one subject
and one direct object’, and that ‘one surface sign of a complex sentence will be the
presence of more than one of these elements’—as phrased by Lawal (: )—was
based on the intuitions of English-speaking theoreticians. The facts of languages with
serial verbs showed otherwise.
Numerous scholars have pointed out the fact that a serial verb construction cannot
be rephrased with coordinated sentences without a meaning change. A single-word
serial verb construction in . in Igbo consists of two verbs, ‘beat’ and ‘kill’, and
means ‘beat to death’ (Lord : ).
. ó ti|-gbù-rù nwóké áhù̩ Igbo
he beat-kill- man that
‘He beat that man to death’ (beat kill)
The two clauses, ‘he hit that man’ and ‘he killed that man’, can be combined; one
clause is marked as ‘consecutive’ to the other in ..
. ó tì-rì nwóké áhù̩ ò̩kpó̩, gbú-é ya| Igbo
he hit- man that blow, kill- he
‘He hit that man and killed him’
In ., the man died as a direct result of being hit. In ., the murderer could have
killed the man by means other than hitting—the man’s death does not have to be the
direct result of being ‘hit’. The consecutive construction in . shows a sequence of
actions, or one action independent of another one (Lord : ). The serial verb
construction in . describes one event—the first component refers to one action,
and the second component refers to the result of that action. Verb sequences in .
and . are not synonymous; deriving one from the other would result in a loss of
meaning. For many serial verb constructions it is hard to come up with an equivalent
consisting of several clauses.7
The tide turned drastically with the publication of Foley and Olson (). This
paper presented the first consistent and cross-linguistically informed line of argu-
ment for the monoclausal analysis of serial verb constructions, their cross-linguistic
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

properties, including argument sharing, and the serializability of verbs. There has
gradually emerged a consensus concerning the characteristics of serial verbs. The
definition in §. reflects this consensus and consolidates the features proposed by
Foley and Olson (); Noonan (); Crowley (, ); Zwicky (); Durie
(); and Jarkey (), among others.8
The treatment of serial verbs as underlyingly conjoined clauses with deleted
subjects was gradually abandoned.9 However, the belief that, historically, all serial
verb constructions derive from a clause union—an erstwhile coordinating or subor-
dinating structure—still lingers. This is what we find in the work by Givón (:
–, : –, : –) and his students. According to Hyman (: ),
serial verbs and verb sequences linked together with a consecutive marker are found
in similar environments in languages with serial verbs, and therefore ‘there seems to
be a natural order from consecutivization to serialization’. The proof for such
developments is mostly lacking; in Chapter  we turn to other mechanisms at work
in the development of serial verb constructions in individual languages. The Appen-
dix to this chapter deals with some of the ways in which serial verb constructions
have been treated by linguists in recent times.
Classification of serial verb constructions is a further matter. The division of serial
verbs into symmetrical and asymmetrical was foreshadowed by Christaller (:
–). What he referred to as ‘combinations of verbs’ were of two kinds—‘essential
combinations’, where ‘one verb is the principal’, and the other one provides a further
specification, and ‘accidental combinations’ where ‘two or more predicates . . . expressing
different successive actions, or a state simultaneous with another state or action, but
having the same subject, are merely joined together’. Sebba (: ) was the first one to
have suggested that serial verbs may consist of a ‘free’ component (from an open class)
and a ‘fixed’ component (recurring member of serial verb constructions, such as hand-
ling or motion, that is, from a closed class). This is a precursor of our division of serial
verbs into symmetrical—consisting of verbs from open classes—and asymmetrical—
consisting of one verb from a closed class (outlined in Aikhenvald a, a, b;
and partly inspired by Durie’s  suggestion to classify serial verbs into ‘balanced’ and
‘unbalanced’).
That verbs within a serial construction may lose their verbal status was aptly
captured by Ansre (): what he called a ‘caveat’ to serial verbs refers to the fact
that in Ewe and Twi (or Akan) some components of serial verb constructions do
not have all the verbal properties and so show signs of grammaticalization. As
Westermann (: ) put it,

. . . many verbs when they stand next to others play the part of English prepositions, adverbs or
conjunctions. Now many of these verbs, in playing the part of prepositions etc., begin to lose
their verbal characteristics, in that they are no longer conjugated; they thus begin to become
form words.
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Serial verbs in linguistic history 

Distinguishing serial verb constructions from other multi-verb structures is not

always straightforward. In many West African languages, serial verb constructions
coexist with consecutive clauses where the second verb is marked with a morpheme
translatable as ‘and’. Consecutive clauses and serial construction can be very similar
in terms of their meaning (see, for instance, the discussions by Hyman  and Van
Leynseele ). And consecutive constructions may be confused with serial verbs.
An example of a consecutive construction from Kana, a Cross River language from
the Benue-Congo family, is in . (Ikoro : –). The connective sā is optional,
but implied if omitted. In . it appears in brackets.
. kúé mɛ̄ (sā) tēērā kìī bē Kana: consecutive construction
he:call me  run go house
‘He called me and then ran home’
A consecutive construction will refer to two separate actions, ‘whereby one is
dependent upon the other’ (Ikoro : ). In contrast, in a serial verb construc-
tion, ‘the temporal ordering of the actions is irrelevant’. The two clauses in consecu-
tive constructions may have different subjects; while in a serial verb, subjects of the
components are always the same.
Along similar lines, the analysis of constructions containing na ‘and’ in Tok Pisin
as serial verb constructions remains problematic (examples are found in Verhaar
). Akkadian consecutive constructions discussed by Kraus () cannot be
considered serial verb constructions for the same reason as Kana consecutive con-
structions: they allow optional inclusion of the conjunction -ma ‘and’.
Sequences of verbs—one of which is marked as syntactically dependent—and even
chains of clauses can have semantic and functional similarities with serial verbs.
Wolaitta, an Omotic language from Ethiopia, has complex verbs with two compo-
nents, one of which is marked as a ‘converb’. These complex verbs can be classified
into symmetrical and asymmetrical, and may express a number of aspectual and
other meanings which serial verbs also express (Amha and Dimmendaal ). But
they are not serial verbs, because one of the components is formally marked as
dependent to the other.
Clause-chaining constructions in Yankunytjatjara, an Australian language, consist
of two verbs; the nonfinal verbs in a sequence take a subordinating suffix and can be
separated from the final verb by a pause. Goddard () misleadingly calls these
‘serial verb constructions’. The term ‘serial verb construction’ in the Tupí-Guaraní
linguistic tradition (e.g. Jensen ) refers to ‘gerund’ constructions composed of
two predicates, one of which is marked as a dependent verb; these do not in fact
qualify as serial verb constructions, as was demonstrated by Seki (). We return to
these issues in Chapter .
The term ‘serial verb’—like any other term in linguistics—has to be applied with
care. As Robert Blust (: ) aptly put it, the ‘coming of age’ of SVCs in the
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

literature ‘works both ways: constructions that may have been overlooked in
grammars written during an earlier era may finally have received further due, but
the “bandwagon” effect of theory-building may also have encouraged the “discovery”
of SVCs where this is only possible by diluting the definition of verb serialization to
the point that it ceases to be distinctive’. This is what we need, and strive, to avoid.

. What this book is about

This book focuses on the grammatical phenomenon of verb serialization world-wide.
What are the parameters of variation of serial verbs? What meanings can they
express? How do serial verbs differ from other, perhaps superficially similar multi-
verb constructions? How do serial verbs emerge, and what happens to them in
language history? What are they good for in terms of representation of event
structure?
Exactly what qualifies as a serial verb construction? Chapter , ‘Recognizing a
serial verb’, addresses the definitional properties of serial verbs—their monoclaus-
ality, prosodic properties, and the overall meaning encompassing one event. Serial
verb constructions tend to share core syntactic arguments—subjects and objects. But
there are also other kinds—including switch-function serial verbs where the object of
the first component is the same as the subject of the second one. No arguments are
shared in just a few types of serial verbs.
In terms of their composition, serial verb constructions divide into asymmetrical
and symmetrical. Chapter , ‘Serial verbs: Their composition and meanings’, con-
trasts the two kinds, outlining the semantic types of each of these.
The components of serial verb constructions can be placed next to each other, or
another constituent may intervene between them. They may form one, or more,
grammatical words. A grammatical category—person, tense, aspect, and others—can
be marked on every member of the construction, or just once per serial verb. In
Chapter , ‘Formal properties of serial verbs’, we look at how these features interact,
and what other properties serial verbs may have and why.
Serial verbs may be a productive technique within a language. Or they may be
restricted to just a few combinations in a limited number of environments. How do
serial verbs differ from superficially similar verb sequences? If a language has serial
verbs, they are always different from other verb sequences, no matter how subtle the
difference. As Bolinger (: ix–x) put it, ‘any word which a language permits to
survive must make its semantic contribution; and . . . the same holds for any con-
struction which is physically distinct from any other construction’. These are the
topics of Chapter , ‘The limits of verb serialization’.
As Durie (: ) put it, within a single language there can be a ‘good case for
distinguishing quite different kinds of serialization’ with different sets of formal and
semantic properties. And indeed, a single language can have more than one kind of
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. What this book is about 

serial verb construction—one may be contiguous, the other not; one consist of a
single word, and the other of more than one. If a language has several types of serial
verb constructions, what meaning differences between them do we expect? This is
what Chapter , ‘The many facets of serial verbs’, is about.
Serial verbs have many grammatical functions. They help dissect a single event
into minute details. The ways in which serial verbs are put together reflects speakers’
experience—what Diller (: ) termed ‘cultural influence in event cohesion’.
These are the topics addressed in Chapter , ‘What are serial verbs good for?’. At the
end of the chapter we discuss common features of languages which have serial verbs.
Chapter , ‘The rise and fall of serial verbs’, focuses on the origins of serial verb
constructions and their loss and gain in different language contact situations. We
discuss grammaticalization and lexicalization of serial verb constructions as major
pathways of their development, and touch upon the ways in which serial verbs are
learnt by children.
The last chapter, ‘The essence of serial verbs’, sums it all up, putting together
conclusions and generalizations made in the previous chapters.
The fieldworker’s guide, ‘Serial verb constructions—how to know more’, contains
a checklist of points to be addressed for an indepth analysis of serial verbs, and
suggestions to fieldworkers working on the phenomenon. It is never enough to just
say that a language has serial verbs and offer a few random examples: to fully
understand the phenomenon, the analyst needs to go in some depth with regard to
the form and the meanings of serial verbs and, if at all possible, their history. The
checklist is not a questionnaire. It aims at offering guidelines for linguistic analysts—
to make sure it is as comprehensive as can be at our present stage of knowledge.
The past few years have seen a surge in new systems of serial verbs discovered and
analysed. It is now time for a thorough cross-linguistic account of serial verbs in their
impressive diversity. I aim at providing a comprehensive, inductively-based analytic
framework for investigating serial verbs, based on a wide range of languages of
different typological profiles and genetic affiliations. This empirical basis will help
ensure that typological generalizations offered here will withstand the test of time,
and be of substantial use to linguists interested in how languages work.
Typology and language analysis feed into each other. In order to come up with
sensible cross-linguistic generalization borne out by the facts of languages, a typologist
needs to rely on good quality comprehensive reference grammars. The converse is also
true. Detailed reference grammars of previously undescribed languages alert typologists
to new phenomena, and offer materials for new typological generalizations.
A typological study is crucial for providing new analytic options and ideas for a
major business of linguists—analysing and writing grammars of unknown or poorly
understood languages. This book is intended to offer scholars a comprehensive
analytic framework, and to alert them to the diversity of patterns in serial verb
constructions—in comparison to other sequences of verbs, encouraging them to
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

undertake fieldwork and see for themselves how languages work, and why they are
the way they are. The analysis is cast in terms of basic linguistic theory—the
fundamental typologically-based theoretical apparatus based on cross-linguistic
inductive analysis of numerous languages.

. The empirical basis, and conventions used

This book is about serial verb constructions, their functions, variation, and history
across the world’s languages. Any inductive study—if it aims at achieving significant
cross-linguistic generalizations and hypotheses—needs to account for as many lin-
guistic systems as possible. This book is based on examination of about  gram-
mars of languages from various language families, linguistic areas and regions of the
world. Following the footsteps of classical typologists, I have made it my task to look
at every language on which I could find reliable information, making sure I cover all
the areas, and as many families as possible. This approach allows me to make the
study as comprehensive as possible in our present level of knowledge of the languages
themselves. My conclusions are truly empirically based.
There is a tendency for some typologists to work with just a sample of languages.
The ways in which such samples are compiled vary a good deal (see Aikhenvald and
Dixon : – for a critique of sampling methodology). This work is not
restricted to a sample of any type.
I could not cite all the examples of every phenomenon—otherwise the book would
have been immense. I usually provide a particularly illustrative example, and men-
tion other similar ones (in an endnote). If a certain phenomenon is found in more
than half of the languages under consideration I call it ‘relatively frequent’; if it is
found in a restricted number of languages (one to ten), I cite all of them and indicate
its rarity. Note, however, that what appears rare to us at the present stage of
knowledge may turn out to be more frequent when we start learning more about
hitherto little known languages and areas. This is the reason why I chose at this stage
not to give any statistical counts. Only about one-tenth of all human languages have
been adequately documented so far—it therefore seems most judicious to follow a
qualitative approach, postponing a proper quantitative analysis to the time in the
future when more data is available and can be scientifically assessed.
A word on conventions. This book contains many examples from—and many
mentions of—languages from various areas and genetic groupings. When the lan-
guage is introduced for the first time, its affiliation and where it is spoken are listed
after its name—for instance, ‘Mupun, a Chadic language spoken in Nigeria’. Later
mentions of the same language do not include this information—which is summar-
ized in the Index of languages at the end of the book.
It is incumbent upon a typologist—or any linguist with common sense—to
separate fact from unsubstantiated hypothesis. Highly speculative entities like
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Appendix 

Indo-Pacific, Amerind, Macro-Equatorial, Arawakan, and the like, have no place in a

piece of scholarly writing. Along similar lines, I eschew unsubstantiated hypothetical
groupings such as Pama-Nyungan in Australia (see the criticism of this hypothesis in
Dixon : – which may have some typological, but no substantiated genetic
relevance), or the ‘Trans-New Guinea phylum’ for non-Austronesian languages of
the New Guinea region. That is, Aboriginal languages of Australia are referred to as
belonging to the Australian area; for Papuan languages a family is given if established,
e.g. Alamblak (Sepik Hill, Papuan area), or Manambu (Ndu, Papuan area).
Language names and language examples are quoted as they are given in the sources
(I have made no attempt to re-spell any of the examples). Language examples are
numbered separately for each chapter, with the first number referring to the chapter
number, e.g. . is the first language example in Chapter , . is second example in
Chapter , and so on. The components of each serial verb construction are in bold.
The literal translation consisting of individual verbs is given after the translation of
each serial verb construction (as we saw in examples .–). The terms ‘serial verb’
and ‘serial verb construction’ (abbreviated as SVC) are used interchangeably.
This book is not an encyclopaedia of serial verbs, nor can it be the last word on
verb serialization and related issues. As Bolinger (: ) puts it, ‘no matter how
wide the net is cast, a fish or two always escapes’. It is my hope that the framework
I offer will be solid enough to secure the catch.

. Appendix: How serial verb constructions have been dealt

with in the linguistic literature
Serial verbs have taken their time in getting into standard paradigms of language
description and analysis. They are conspicuously absent from prescriptive grammar-
writing guidelines such as Derbyshire and Pullum’s Handbook of Amazonian languages
and Comrie and Smith’s Lingua descriptive series questionnaire. The framework of
this questionnaire forced Faraclas (: –), in his comprehensive grammar
of Nigerian Pidgin, to discuss serial verbs within the chapter on coordination where
they do not really belong, simply because, as is obvious from the rest of the grammar,
they are not equivalent to coordinating structures. The discussion of serial verbs in
Ndyuka, a Suriname Creole, is split between syntax (Huttar and Huttar : –)
and verbal morphology (pp. –). Payne’s introduction to grammar writing (:
–) offers a brief discussion of serial verbs without a clear definition and no
statement of how to distinguish them from other verb sequences. A recent guide for
writing grammars (Aikhenvald : –) contains defining features of serial verbs
and how to tell them apart from other, perhaps superficially similar structures.
The definition of serial verbs in many languages has been marred with a transla-
tional bias. Scholars influenced by familiar Indo-European and Semitic categories
consider serial verbs used in their function of complementation strategies as
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

containing zero infinitive markers (see, for instance, Owens and Khamis : ;
and the criticism of early Chinese grammarians by Matthews : –); the term
‘pseudocomplementation’ suggested by Seuren () is equally misleading.
Difficulties which arise with respect to a definition of what an SVC is—and what it
is not—have led some scholars to deny the very existence and cross-linguistic
importance of this phenomenon, especially within an analysis limited by a particular
formalism (e.g. Law and Veenstra  and Paul , with a special relevance to
Chinese). Depending on how criteria are applied, what is considered serial verbs by
one author may not be regarded as such by another (for instance, some of the serial
verb constructions in Mandarin discussed by Li and Thompson (: –) are
not treated as such by Matthews (: ) in his study of Cantonese.
A comprehensive review of various formal approaches to serial verb constructions
(e.g. Baker ; Déchaine ; Stewart ; and references there) and their pitfalls
would require a book-length study. Just a few remarks are in order. The requirement
of obligatory object sharing and a ban on duplicate roles in serial verb constructions
suggested by Baker () are not borne out by a close analysis of individual
languages. (This requirement is no doubt rooted in his theoretical stance, which
demands postulating structural equivalence between a simple verb in a language like
English and serial verb constructions in serializing languages, in terms of argument
structure and the like.) His statement concerning ‘double headedness’ of serial verb
constructions is equally dubious. The ‘headedness’ of serial verb constructions
depends on their types: only asymmetrical SVCs have clear syntactic heads. Durie
() offers a concise critique of Baker’s approach.
The oft-quoted ‘serialization parameter’ (Stewart ) provides an investigation
of just a few properties of serial verb constructions in a West African language (Èdó),
limiting SVCs to just two types (‘resultative’ and ‘consequential’). A number of
conclusions, for example, that ‘no verb in the serial verb construction can bear
morphological tense inflection’ (Stewart : ) are not borne out by the facts
of the world’s languages. Neither are the lexical constraints and argument-sharing
properties (the topic of §.. of this book): for instance, Stewart claims that only in
‘consequential’ serial verb constructions are there ‘sequences of two transitive verbs’.
We will see in Chapter  that this is not true. To be able to successfully formulate a
cross-linguistically valid ‘serial verb construction parameter’ which would explain
why some languages have SVCs and others do not, one needs a broader perspective
on cross-linguistic variation in verb serialization.
A number of early attempts to classify serial verbs focused on some of their
features. For Yoruba, Bamgbos̩e () distinguished between ‘linking’ (or sequen-
tial) and ‘modifying’ (or manner) serial verbs. In an attempt to reflect different
relations between the components, Schiller (b) divided them into coordinating
serial verbs (which have to share subjects) and subordinating, or modifying serial
verbs where one verb provides manner modification to the other. The terms
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Appendix 

‘coordinating’ and ‘subordinating’—when used to describe serial verbs—obscure the

internal structure of SVCs which are neither coordinating nor subordinating. Lawal
() adds to that a third type—complex serial verb constructions which involve
decomposition of a concept rendered by one serial verb and which would be rendered
by separate verbs in non-serializing language.
Veenstra (: –) classified serial verbs into ‘clausal serial verb constructions,
exhibiting relatively more independence between the different subevents denoted by
the separate verbs, and free lexical selection’ and ‘phrasal serial verb constructions,
exhibiting relatively less independence . . . and a limited set of participating verbs’ (see
also Adone : ). Veenstra himself acknowledges that ‘the notions clausal and
phrasal are not clearly defined and do not (necessarily) have implications for the
(structural) analysis of serial verb constructions. He suggests two pairs of additional
parameters: lexically restricted versus lexically free, and less independence between
subevents versus more independence between subevents (yielding four types). The
parameter ‘lexically restricted’ or ‘lexically free’ reflects restrictions on the type of
verb in an SVC, reminiscent of the asymmetrical versus symmetrical constructions,
but less clearly defined. Using Veenstra’s words, ‘the rather vague notion’ of ‘more or
less independent subevents’ reflects ‘event composition’ of an SVC: ‘more inde-
pendent subevents’ involve any event predicates, and ‘less independent subevents’
involve modality, state, event, and a combination of state predicates (p. ). The
application of these notions is not self-explanatory, especially since degree of
‘dependency’ between subevents is determined by intuition only.
Within the framework of Role and Reference grammar, the distinction between
different types of serial verbs is described as ‘core’ versus ‘nuclear’ serialization (see,
for instance, Foley and Olson ; Crowley ; Jauncey : –). This is
based on a distinction between ‘core’ and ‘nucleus’ of the clause and an arbitrary split
of grammatical categories into core and nuclear inherent to that theory. This
terminology reflects the hypothetical nature and the degree of syntactic juncture
between the verbs which form a serial construction. The basic difference between the
two is that ‘while core layer serialization allows some degree of independence to
the two verbs in the choice of nominal arguments associated with each, this is not
the case with nuclear serialization’ (Crowley : ). The applicability of such a
binary classification is problematic. First, it is based on a layered hierarchical view of
clausal structure into ‘nucleus’, ‘core’, and ‘periphery’, and an arbitrary classification
of some categories such as tense as ‘peripheral’ and others such as aspect as ‘nuclear’
(see the critique of Foley and Van Valin’s () classification of verbal categories by
Reid : , based on the facts of Ngan.gityemerri where tense is marked closer
to the verbal root than aspect and would thus be best considered more ‘nuclear’ than
aspect). Secondly, different semantic and structural types of serial verb constructions
within a language may display different restrictions, going beyond a binary division.
Thirdly, the distinction between ‘core’ and ‘nucleus’ is largely based on the intuitive
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . The framework

notion of ‘independence’ of components, without taking account of the mechanisms

in putting verbs together and deriving the ensuing meanings. Approaching the
diversity of coexisting serial verbs in terms of their formal and semantic features
(rather than trying to fit them into arbitrarily defined core and nuclear categories)
allows for a more fine-tuned understanding of the mechanisms at play. The notions
of ‘core’ and ‘nuclear’ are not applied consistently across grammars. In many
grammars of Oceanic languages, serial verb constructions with single marking are
considered ‘nuclear’ no matter what their semantics and principles of argument
sharing; and those with concordant marking are considered ‘core’ (see, for instance,
Hyslop : ).
Just about every year sees a reappraisal of what serial verbs are. Bisang ()
suggests replacing all the grammatical criteria for serial verbs just by the semantic
one, claiming that their monoclausality and other features can be ultimately derived
from serial verbs representing a single ‘event’ (defined in a vague way). The con-
structions which Bisang deems as representing more than one event (such as ‘multi-
scene’ serial verbs in Kalam) are arbitrarily considered by him as sequences of
clauses, thus going against the language analysts (e.g. Pawley and Lane ). In an
attempt to redefine serial verbs, Haspelmath () defines a serial verb construction
as ‘a monoclausal construction consisting of multiple independent verbs with no
element linking them and with no predicate–argument relation between the verbs’.
This definition is fairly problematic. It arbitrarily includes only those constructions
which consist of ‘independent verbs’—presumably, only multi-word constructions,
excluding productive single-word constructions found in numerous languages. The
obscure formulation ‘predicate–argument relation between the verbs’ arbitrarily
excludes serialization of complement-clause taking predicates and causative serial
verb constructions (widespread types of serial verbs in many languages, as we will see
in Chapter ). Sharing tense, aspect, modality, and mood are deemed to be ‘unneces-
sary’ criteria for serial verb constructions (and yet they are used, by the author, to
demonstrate monoclausality of serial verbs). Semantic labels of ‘agent’ and ‘patient’
are confused with syntactic labels of ‘subject’ and ‘object’. There are, in addition, a
fair number of factual errors (see Aikhenvald and Dixon forthcoming for a critique).

Notes
. Such combinations of the two motion verbs come and go with another uninflected verb have
been named ‘quasi-serial verb constructions’ by Pullum (). We return to these in §..
. The term ‘compound’ is used in Chinese linguistic tradition, to refer to contiguous serial
verb constructions in Chinese varieties which tend to form one grammatical word and
usually have a resultative or cause-effect meaning (for instance, Tham ). Their status as
serial verbs was demonstrated by Wu () and Matthews (: ). Along similar lines,
Kiessling (: –) treats single-word symmetrical serial verbs in Isu, a Grassfields Bantu
language, as compounds.
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

Notes 

. Combinations of verbs were described as ‘connection of sentences without any conjunction’

for Akan by Riis (: –), and as ‘verbal separable compounds’ and ‘combinations of
verbs with verbs’ for Ga, another Kwa language, by Zimmermann (: –, –).
. The original text reads: ‘Die Vorstellung mehrerer Geschehnisse, meistens nur zweier,
können zu einer neuen Vorstellung zusammengefasst werden, ähnlich wie im Deutschen
durch Vorsilben Vorstellungen präzisiert werden in “weichen, ausweichen” . . . Dazu werden
volle Verbalformen hintereinander gestellt’ (translation from Dempwolff et al. : ).
Along similar lines, Ward (: ) points out that ‘what in English may be expressed by
one verb consists frequently of a series of actions: in Yoruba, as in other West African
languages, the separate actions into which the idea may be analyzed, are usually expressed
by separate verbs. For example, the English word bring implies the action of taking
(something) and coming; fetch implies going, taking and coming. These must be expressed
in Yoruba.’
. The term used by Stewart () is ‘serial verbal sentence’ (see also George : ). Twi is a
dialect of Akan. In his recent attempt to re-evaluate the notion of serial verbs, Haspelmath
(: ) erroneously suggests that the term ‘serial verb construction’ was coined by
Stewart ().
. See, among others, the pioneering work by Williamson (); Ansre (); Boadi ();
Matisoff (, ); Stahlke (); Williams (); Awobuluyi (); Lord (,
); Bamgbos̩e (); Schachter (); and George (). An annotated bibliography
of serial verb constructions is in Aikhenvald (forthcoming); see also Sebba (: –).
. In his attempt at a transformational analysis of serial verbs, Stahlke () noted that a
serial verb construction in Yoruba cannot be paraphrased with coordinated verbs if one
wishes to preserve the same meaning—we will find examples in .–, and the discussion
in §.. There followed a number of attempts to derive serial verb constructions from an
underlying simple sentence. There is further discussion in Van Leynseele (), Foley and
Olson (), and also Lawal (). Schachter () acknowledged the existence of
adverbial SVCs (which do not necessarily ‘derive’ form conjoined clauses).
. See also papers in Joseph and Zwicky () and Lefebvre ().
. For instance, Jansen, Koopman, and Muysken () analyse serial verbs as forming one
verbal phrase, as do later studies, such as Li and Thompson () and Sebba (). Early
attempts at deriving serial verb constructions from underlying complex sentences involved
complicated rules for shared argument reduction and conjunction reduction, cast within
the formal theory of the day (including the Minimalist framework and many more,
particularly notable in the studies of serial verbs in Creole languages, e.g. Veenstra ).
Sadly, even in recent publications, serial verb constructions occasionally continue to be
described as ‘linked clauses’ which ‘behave like a sequence of verb phrases’—e.g. Watters
(: ). However, within the same volume, Creissels (: ) provides arguments in
favour of a monoclausal analysis of serial verb constructions.
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

Recognizing a serial verb

A serial verb construction is a sequence of verbs which act together as a single

predicate. This is what we start with in §.. Serial verbs are always monoclausal—
the topic of §.. Prosodic properties of serial verb constructions distinguish them
from sequences of clauses—we turn to this in §.. The components of a serial verb
construction share tense, aspect, modality, reality status, evidentiality, mood, and also
polarity values—the topic of §.. A serial verb construction typically refers to what can
be conceptualized as one event. This is discussed in §.. In §. we turn to how serial
verbs share subjects and other arguments. The last section offers a summary.

. Serial verb as one predicate

A serial verb functions on a par with monoverbal clauses in discourse, and occupies
one core functional slot—that of a predicate. Verbs which form a serial verb
construction act together as a syntactic whole. A serial verb construction may be
translated with one single verb into a non-serializing language—this is what we see in
many of the English translations in the examples that follow.
The components of a serial verb construction cannnot take separate markers
of syntactic dependency. In Kambera, an Austronesian language from Indonesia,
if a serial verb is the predicate of a relative clause it takes one relativizer per
construction—shown in . (Klamer : ). Dependent clauses are in square
brackets.
. na pulung jia-ya na [pa-laku Kambera
 word -sgA  .-go
ngàndi-na]
take-sg
‘The gospel (lit. word) is what he brought’ (go-take)
In ., from Toqabaqita, an Oceanic language spoken on the Solomon Islands, the
relative clause marker na occurs once per serial verb characterizing it as a whole
(Lichtenberk : ). The serial verb construction ‘work be together’ has a
sociative meaning ‘work with’ (see §..).

Serial Verbs. First edition. Alexandra Y. Aikhenvald.

© Alexandra Y. Aikhenvald . First published  by Oxford University Press.
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Serial verb as one predicate 

. ka faqarongo-a toqa nia ki [na kere Toqabaqita

sg: tell-: people sg   pl:
raa ofu bii nia] . . .
work be.together  sg
‘ . . . he told those people of his who were working with him [to do such and
such]’ (work be.together)
If a serial verb occurs within a subordinate or an embedded clause, its components
cannot be embedded independently. A serial verb construction within a non-main
clause in Tariana takes one subordinating morpheme, just like one verb would. This
is illustrated with . where ‘see’ occurs with a complement clause ‘that I took it
across’ containing a serial verb ‘take make cross’.
. [nhuta nu-thaketa]-ka di-ka-pidana Tariana
sg+take sg-cross+- sgnf-see-..
‘He saw that I took (it) across’ (take cross)
Along similar lines, a serial verb construction in a purpose clause in Sri Lanka Malay
takes one marker of syntactic dependency, mà-, confirming its status as one unit
(Nordhoff : ).
. Aajuth=yang buurung mà-angkath baapi su-diyath Sri Lanka Malay
dwarf= bird -lift take.away -try
‘The bird tried to carry the dwarf away’ (lift take away)
If a serial verb is nominalized, there is one marker of nominalization per construc-
tion. Example ., from Toqabaqita (Lichtenberk : ) shows that the whole
serial verb kwaqe fole ‘split’, literally ‘hit split’, is within the scope of the nominalizing
suffix. The suffix is attached to the last component of the serial verb. The nominalized
serial verb is in square brackets.
. [kwaqe fole-la-na] niu qe aqi Toqabaqita
hit split--sg: coconut sg: .
si qafetaqi
sg: be.difficult
‘Splitting coconuts (e.g., using an axe) is not difficult’ (hit split)
A nominalized serial verb construction in ., from Tariana, also takes one marker
per construction. Note that the subordinator -ka goes at the end of the serial verb
in ., while the nominalizer (in this case, -ri) appears on its first component.
. phe-ri pa: Tariana
+enter- +go
‘entry, entrance’ (lit. the process or the place of going inside) (enter go)1
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . Recognizing a serial verb

Alternatively, every component of an SVC can take the same affix, as in Lango
(Nilotic). In ., both components of an embedded SVC are in the purposive form
(termed ‘infinitive’ by Noonan : ).
. ámìttò [cwɛ̀ kàttò rwòt] Lango
sg+want+ be.fat+ exceed+ king
‘I want to be fatter than the king’ (be fat exceed)
The monopredicative status of serial verb constructions can be corroborated by
answers to a yes-no question. In Lao (Enfield a: ), a ‘yes’ answer to a question
which includes a single-word serial verb has to contain the whole construction. This
is shown in ..
Question:
.a man2 nii3-paq2 naang2 qan0-nii4 bòò3 Lao
sg flee-abandon young.woman -. .
‘Did he abandon (flee abandon) that young woman?’
Answer:
.b nii3-paq2 Lao
flee-abandon
‘(Yes, he) abandoned (her)’ (flee abandon)
Similarly, in Tariana an answer to a question containing a single verb predicate involves
the repetition of just this verb. An answer to a question containing an SVC involves
repeating a whole construction, and never just one component (see Aikhenvald a).
In Goemai (Hellwig : ), serial verb constructions behave as single predicates
with regard to the placement of back-channelling interjections such as mh ‘yes’ and
kwai ‘no’. Addressees back-channel after a clause. If a clause contains a serial verb, an
interjection has to follow the whole construction, and cannot interrupt it:
. N: la goe mang goe su goe wa Goemai
 sgm take sgm run(sg) sgm return(sg)
n-ni / A: mh
-sg.. 
‘If you take (her and) you run (and) you return with her,’ ‘Yes’
(take run return)
N: nk’ong b’it vel / b’ep mûaan tal A: yes
back day two return go(sg) greet 
‘after two days, (she) returns (and) goes (and) greets’ ‘Yes’
(return go greet)
Further, language-specific, features may point towards the monopredicative charac-
ter of serial verbs. Tariana employs repetition to link clauses within discourse. Then, the
whole serial verb has to be repeated, and not just any of its components. Example .
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Serial verbs as one clause 

comes from a narrative about travelling back home from a rubber plantation and then
going on to another location. In the first clause, the speakers came back (this is expressed
with a serial verb, return-arrive). The second clause partly repeats the first clause: the
tense-evidentiality markers are omitted, and the whole serial verb is repeated. This type
of repetition is known as ‘head-tail linkage’, or bridging repetition (Dixon ). The
predicate of the second clause bears linking intonation which involves rising pitch on
the last syllable. This clause is followed by another full clause, indicating what happened
next. Clauses are in square brackets, with their type in subscript.
. [te wa-dia waka-na]main clause Tariana
up.to.that.point pl-return pl+arrive-..
[nese wa-dia waka]bridging clause (rising intonation)
then pl-return pl+arrive
[wa:-na-pita nese-aya]main clause
pl+go-..- there-
‘At that point we arrived back, then as we arrived back, we went right there’
Monopredicative reading of serial verb constructions may be spontaneously corrob-
orated by the intuitions of native speakers. We can recall, from §., that a serial verb
is often best translated with a monoverbal clause into a language with no serial verbs.
Serial constructions never contain a dependent or a nominalized form of any of their
components. For this reason, complex verb forms like perfect or continuous in English
are not serial verb constructions (see further arguments in Zwicky : ). Nor are
converb constructions, or clause chains with a dependent verb. How to distinguish serial
verb constructions from other kinds of verb sequences is the topic of §..

. Serial verbs as one clause

Serial verb constructions form one clause. Monoclausality of serial verbs follows from
their monopredicative character.2 Serial verb constructions allow no markers of
syntactic dependency on any of their components (as we saw in Chapter ). This is
criterial in distinguishing serial verbs from coordination, consecutivization, comple-
ment clauses, subordinate clauses, and other multiclausal structures. The presence of
an overt linker—expressed with a conjunction, as in Nupe, or with a change in tone,
as in Igbo—helps distinguish consecutive constructions from serial verbs in many
African languages (Watters : –).3 More on this in Chapter .
Serial verb constructions may constitute one grammatical word. Single-word serial
verbs are monoclausal. This is the case in Papuan languages of the Sepik (including
Alamblak (.), Manambu, and Yimas) and Igbo (Benue-Kwa: see Lord , ,
and .a).4
Trying to break down a serial verb construction into a set of independent clauses in
an attempt to ‘derive’ a serial verb from an ‘underlying’ clause sequence is not a
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . Recognizing a serial verb

productive way to go. Coordinate structures which contain the same verbs as a serial
verb construction will differ in meaning from SVCs (see Foley and Olson : –,
and examples .– in Chapter ). The serial verb construction ‘take come’ in Yoruba
means ‘bring’, as in . (Stahlke : ).
. mo bá e̹ mú ìwé wá ilé Yoruba: a serial verb
I for you took book came home
‘I brought a book home for you’ (take come)
A sequence of coordinated clauses, one containing the verb ‘take’, the other one
containing the verb ‘come’, has a different meaning—it refers to a sequence of
independent actions, ‘taking’ and ‘coming’. . does not contain a serial verb.
. mo mú ìwé, mo sì wá ilé Yoruba: coordinated clauses
I took book I and came home
‘I picked up a book and came home’
As Stahlke (: ) puts it, ‘it would be possible to continue a coordinate structure’
like . with . (which contains a serial verb).
. s̩ùgbo̩n
́ mo gbàgbé láti mú wá pèlú Yoruba: serial verb
but I forgot to take come with
‘But I forgot to bring (it) along’ (take come)
To continue a serial verb in . with . ‘in the same way would render it
nonsensical’: . states that I did bring the book, and . provides contradictory
information.
An attempt to rephrase a serial verb as a set of coordinate clauses can produce
striking results. ., from Anyi-Sanvi, a Kwa language, contains a serial verb
construction ‘catch eat’ (Van Leynseele : –). The two verbs share subjects
and objects (we return to this in §.).
. cù̩á cì ̩ ákɔ́ !
dì Anyi-Sanvi: a serial verb
dog catch+ chicken eat
‘The dog eats a chicken’ (catch eat)
If each verb is used in an independent clause and the two clauses are coordinated, the
sentence has a totally different meaning. Example . contains two coordinated
clauses. Note the separate subject marking on the second verb ‘eat’ which then has
the meaning ‘copulate’. Unlike in the serial verb construction in ., each verb has
its own subject and also object, so arguments are not shared.
. cù̩á cì ̩ ákɔ́ ò-dì í̩ Anyi-Sanvi: coordinate clauses
dog catch+ chicken he-eat it
‘The dog catches a chicken and copulates with it’
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Serial verbs as one clause 

Paraphrasing a serial verb with two clauses may result in an ungrammatical, or a

semantically bizarre sentence. .a, from Igbo, is a single-word contiguous serial
verb construction (Lord : ).
.a ó tì-wà-rà étéré à Igbo: a serial verb
he hit-split.open- plate the
‘He shattered the plate’ (hit split.open)
This cannot be paraphrased with a sequence of sentences like .b and .c, each
containing one of the components of the serial verb in .a. The sentence in .b is
semantically odd, while .a is not: as Lord (: ) puts it, ‘only a lunatic would
try to beat a plate’; in contrast, to ‘shatter’ a plate, as in .a, is perfectly normal.5
.b *ʔó tì-rì étéré à Igbo
he hit- plate the
?’He hit the plate’
.c *ó̩ wà-rà étéré à
he split.open- plate the
Example .c is unacceptable for a different reason: the verb ‘split open’ on its own
is intransitive, and cannot take an object. In Igbo, a verb can have different transi-
tivity and argument structures when used within a serial verb construction and when
used on its own (Lord : –, –; also see .– in §.).
Even if a serial verb can be paraphrased with two clauses, there is always some
semantic or pragmatic difference. The SVC in ., from Taba, repeated from .,
describes one event: the death of a pig comes ‘as a direct and immediate consequence
of the pig’s being bitten’ (Bowden : –).
. n=babas welik n=mot do Taba: a serial verb
sg=bite pig sg=die 
‘It bit the pig dead’ (bite-die)
The same verbs as in . occur in ., but as coordinated predicates. The meaning
is subtly different: in . ‘there may have been a considerable time elapsed between
the biting and the pig’s eventual death by bleeding’; that is, the death of the pig could
have occurred as an indirect consequence of having been bitten, but did not have to
occur. This is in contrast to . where the biting and the dying are very close.
. n=babas welik n=ha-mot i Taba: coordinated clauses
sg=bite pig sg=-die sg
‘It bit the pig and killed it’
The scope of adverbs, or clausal modifiers, may differentiate serial verbs from
coordinate or subordinate structures. A number of adverbs in Yoruba precede the
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . Recognizing a serial verb

serial verb as a whole, providing modification to the whole construction. This is

shown in .a and .a (see Lawal : –).
The adverbs tètè ‘quickly’ and tún ‘again’ cannot be moved to precede the second
component and modify it. This is what makes examples .b and .b ungram-
matical: the adverb has the whole serial verb within its scope, and not its individual
components.6

.a Olú tètè jókòó kawé Yoruba

Olu quickly sat read
‘Olu quickly sat down and read’ (sit.down read)

.b *Olú jókòó tètè kawé Yoruba

Olu sat quickly read
?‘Olu sat down and quickly read’ (sit.down read)—???

.a O tún mu otí yo Yoruba

he again drank wine full
‘He is drunk again’ (drink full)

.b *O mu otí tún yo Yoruba

he drank wine again full
‘He is drunk full again’—???

Reflexivization is another criterion in favour of Lawal’s (: ) conclusion that

serial verbs in Yoruba are ‘like simple sentences which contain complex verbs’. An
example of a reflexive in a monoverbal clause is in ..
. Mo we araàmi Yoruba
I bathe myself
‘I bathe myself ’
A reflexive pronoun in the language cannot have an antecedent in a different clause
or sentence. Following this rule, using a reflexive in a subordinate clause in .
results in an ungrammatical sentence.

. *Mo so pé Femi we aràmi Yoruba

I said that Femi bathe myself
???‘I said that Femi bathed myself ’

Reflexives apply to a serial verb construction in exactly the same way they would
apply within a single clause, as in . and .. Having a non-reflexive pronoun
(‘me’ rather than ‘myself ’) in either of these would be ungrammatical.
. Mo se ju araàmi Yoruba
I do pass myself
‘I surpassed myself (do pass)’
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Prosodic properties of serial verbs 

. Won fa araàwon wá Yoruba

they drag themselves came
‘They dragged themselves here’ (drag come)
Causative constructions behave differently. A reflexive pronoun cannot be used in a
multi-verb causative construction shown in .a.
.a Mo jé Bólá ran mi lówó Yoruba
I made Bola help me in.hand
‘I made Bola help me’
.b is ungrammatical (Lawal : ).
.b *Mo jé Bólá ran araàmi lówó Yoruba
I made Bola help me in.hand
???‘I made Bola help myself ’???
This indicates that a causative construction consists of two clauses, and is not a serial
verb construction (see also §..).
We return to other features such as clefting and focusing of components of serial
verbs as a way of differentiating them from coordinate clauses in §... Further
support for the monoclausality of serial verbs comes from their prosodic properties.

. Prosodic properties of serial verbs

A serial verb construction sounds like a monoverbal clause, and not like a sequence of
clauses. In many languages clause boundaries are indicated by an intonation break;
no such intonation break or pause markers can occur between the components of an
SVC. The lack of pause or intonation break between sequences of verbs within a
multi-word serial verb construction differentiates it from a sequence of verbs, each in
an independent clause. ., from the Lolovoli dialect of north-east Ambae, an
Oceanic language from Vanuatu, is an example of a multi-word serial verb construc-
tion. It is characterized by the absence of intonation break, and also shared marking
of aspect and reality status. We return to these in §..
. Gai-rue ra=ru mo vano ra=ru mo rivu talu Ambae
-two nsgS=du  go nsg=du  plant garden
‘Two went to plant the garden’ (go plant)
In contrast, . is a sequence of clauses, with an intonation break between two verbs
(indicated by a comma) (and no aspect and reality status sharing). Clauses are in
square brackets.
. [Ra=u hivo,] [ra=mo rivu butete] Ambae
nsg.S= go nsgS= plant sweet.potato
‘They have gone down to the garden, and they’re planting sweet potato’
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . Recognizing a serial verb

Intonation is often the only means of distinguishing serial verbs from juxtaposition
of clauses in the Lolovoli dialect of the north-east Ambae language, and also in Taba
(see Bowden : –). Similarly, in Dyirbal, Toqabaqita, and Tariana, serial
verbs are pronounced as one intonation group, with no pause between the compo-
nents (Lichtenberk : ; Aikhenvald b; Dixon : –). Components
of serial verb constructions in Thai are ‘uttered with single breath-group and
intonation contour; repair properties suggest a unified articulatory assemblage’
(Diller : ). Speech ‘repair’ in serial verbs constitutes a further piece of
evidence in favour of serial verbs behaving like a single predicate. If a speaker of
Tariana makes a mistake and falters in the middle of a long serial verb, they start the
whole construction all over again—in the same way as they would do with a longish
monoverbal predicate (Aikhenvald b: ).7

. Shared tense, aspect, modality, reality status, evidentiality,

mood, and polarity values
Individual components of a serial verb construction share tense, aspect, modality,
reality status, evidentiality, mood, and also polarity values. This implies that no
independent choice or contrast in any of these categories should be possible for the
individual components of a serial verb. This makes them all the more different from
multiclausal structures. Shared marking of aspect and of realis in the Lolovoli dialect
of north-east Ambae (Hyslop : ) helps distinguish a serial verb construction
from a sequence of juxtaposed clauses, as we saw in .. Each category may be
marked once per construction (‘single’ marking), or on each component (‘concord-
ant’ marking): it is the value of the category, and not the details of its marking, that
make a serial verb a single unit. We return to this in §..
Sharing the value of a verbal category—such as tense, aspect, and others—may imply
that the components of a serial verb must have the same marking of that category. The
category itself can be marked on all the components, as we saw in the previous examples.
In Lango, both components of an SVC have to take the same marking for all verbal
categories (Noonan : –). Both verbs are marked as ‘habitual’ in ..
. àcwɛ̀ àlɔ̂ rwòt Lango
sg+fat+ sg+exceed+ king
‘I am fatter than the king’ (I-fat I-exceed)
In Dyirbal (Dixon : , ), the components of a serial verb construction agree
in their inflection. In a multi-word serial verb construction in . both verbs take the
past tense inflection.
. ba-ŋgu-l nudi-n yugu jayŋu-n Dyirbal
there--masc cut- tree finish.off-
‘He finished chopping down the trees’ (cut finish)
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Shared tense, aspect, and other categories 

Similar examples can be found in Yidiñ (Dixon : –), and a number of other
Australian languages (see, for instance, Reid  on Ngan.gityemerri, Green 
on Gurr-goni). Sharing of tense and mood was identified as a definitional property of
serial verbs in Ewe by Westermann (: ; quoted in §.).
Even if the language does not have obligatory tense marking, the components of a
serial verb must have one temporal reference. In Fongbe (Lefebvre and Brousseau
: ), one verb in an SVC cannot describe an action happening at one time, and
the other one at another time. . is ungrammatical:
. *Kɔ̀kú sɔ́ sákì ɔ́ gànɖókpómɛ̀ yì àxì mɛ̀ gànwèmɛ̀ Fongbe
Koku take bag  one.o’clock go market in two.o’clock
Alternatively, a marker of aspect or tense can occur once per serial verb
construction, characterizing it as a whole. In Cantonese, the marker of experiential
aspect gwo3 typically occurs on the second component, as in . (Matthews
: ).
. lei5 gan1 jan4dei6 hok6-gwo3 Zung1man2 Cantonese
you follow people learn- Chinese
‘You have learnt Chinese from someone’
Along similar lines, a serial verb construction with an instrumental meaning in
Kristang (Baxter : ) contains just one marker of perfective aspect (see also
. in §.. on reducing valency).
. eli ja tomá faka kotrá kandri Kristang
sg  take knife cut meat
‘He cut meat with a knife’
A serial verb construction in Tariana will take one marker of evidentiality (gram-
maticalized information source) fused with tense. The marker attaches to the last
component, as shown in . (and also §.).
. nu-dia nuká nu-wa nu-a-na Tariana
sg-return sg+arrive sg-enter.the.jungle sg-go-..
‘I managed to return going into the jungle away from here’ (return-arrive-
enter the jungle-go)
Serial verb constructions in Dyirbal and Lango exemplify concordant marking of
categories. Single marking is a feature of Cantonese, Kristang, and Tariana.
Marking the same value of mood and modality can be distributed between the
verbs, following strict rules. In Paamese, an Oceanic language, if the first verb in an
SVC is marked for prohibitive, the subsequent verbs acquire potential marking
(Crowley : –; : ), as in .. The whole serial verb is within the
scope of negation.
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . Recognizing a serial verb

. mu-na-sooni-tei auve na-hinaa Paamese

pl--throw- throwing.stick sg.-go.up
‘Don’t throw the throwing stick up’ (don’t throw go.up)
The whole construction has a prohibitive meaning. In commands couched as
serial verb constructions, the first verb takes the imperative form, and the subsequent
verbs are marked for ‘immediate’ or ‘distant’ mood. An example is at . (a similar
principle has been described by Green : , for Gurr-goni) (see also §..).
. saani-e vaa-mai Paamese
sg..give-sg sg..-come
‘Give it to me’ (i.e. give it hither) (give come)
Markers of illocutionary force typically characterize the serial verb construction as a
whole. The modal expression wasã ‘let’s!’ in Tariana has scope over the whole
construction, as shown in ..
. deri-nuku wasã wheta wa-hña Tariana
banana-../ let’s pl+take pl-eat
‘Let’s get a banana!’ (take eat)
Similarly, adverbial lexemes in Emmi, an Australian language (Ford a: –),
have scope over both verbs within a serial verb construction. An adverbial lexeme can
precede or follow the serial verb construction, and has scope over both components.
A two-verb serial verb construction with the adverbial phrase ‘tomorrow’ (literally,
for next day) is illustrated in . (Ford a: ).
. malh ngave+rr=no ngadi Emmi
story ...pick.up+tell.story= ...sit
yerranguya=no
next.day=
‘I’ll tell you a story tomorrow’ (tell.story sit)
A serial verb construction is typically negated as one whole. In Toqabaquita, the negative
subject marker and the negative verb precede the verb complex, and has the whole serial
verb construction within its scope (Lichtenberk : ), as shown in ..8

. qe aqi qosi qini feto-a Toqabaqita

sg: . sg: pinch pinch.and.twist-:
wela qena!
child that
‘Don’t pinch the child!’ (To inflict more pain, the victim’s skin is pinched and
twisted. This is the usual way of pinching someone) (pinch pinch.and.twist)

One negator will have the whole construction as its scope, as in . (Noonan : ).
 OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

. Shared tense, aspect, and other categories 

. pé àcwɛ̀ àlɔ̂ rwòt Lango

 sg+fat+ sg+exceed+ king
‘I am not fatter than the king’ (not I-fat I-exceed king)
An inherently negative verb in a serial verb construction makes the whole construc-
tion negative. Example . from Mav’ea, an Oceanic language from Vanuatu,
contains a negative verb leng ‘be unable to, cannot’ as V (Guérin : ):
. na-leng on rongo=a Mav’ea
sg-cannot look feel=sg
‘(This woman,) I have never seen her’
In Mupun, a Chadic language from Nigeria (Frajzyngier : ), the discontinu-
ous negation marker bà . . . kàs encompasses the serial verb construction and its
arguments, as in .a.
.a ba wu se siam n-pankshin kas Mupun
 masc depart go.down -Pankshin 
‘He didn’t go down to Pankshin’ (not depart-go.down)
If a negator is inserted between two verbs they refer to two different events and are
understood as coordinated, as in .b.
.b wu ji ba wu se siam Mupun: coordinated clauses
masc come  masc depart go.down
n-pankshin kas
-Pankshin 
‘He came [here] and/but he didn’t go down to Pankshin’
Alternatively, components of an SVC have to take the same negative marking (we return
to concordant marking of linguistic categories in serial verbs in §..). An example of
such concordant marking from Anyi-Sanvi is in . (Van Leynseele : –).

. cù̩á ńjî ̻ ákɔ́ ń+Iní Anyi-Sanvi

dog +catch+ chicken +eat+
‘The dog never eats a chicken’ (not-catch not-eat)

In contrast, a coordinate structure—containing the same verbs—can contain a

negative and a positive verb:

. cù̩á cì̩ ákɔ́ óŋgṹ ì ̩ Anyi-Sanvi: coordinate clauses

dog catch+ chicken he++kill+ it
‘The dog catches a chicken and does not kill it’

The scope and the options for negation may further differentiate between serial verbs
and other multi-verb constructions (see examples from Fongbe in Lefebvre and
OUP CORRECTED PROOF – FINAL, 20/9/2018, SPi

 . Recognizing a serial verb

Brousseau : –, and §.; Table  in Diller  shows how serial verbs in
Thai differ from other superficially similar multi-verb structures in their shared
polarity, modality, and temporal setting).
Since components of a serial verb construction cannot be negated separately,
ambiguity may arise. In Alamblak, only one negative word can occur with a single-
word SVC. The scope of negation can be the whole construction, or any one of its
components by itself, or any combination of contiguous components. A negated SVC
may then be ambiguous in several ways. This is shown in ., with a serial verb
construction ‘roast-get-go’. Each of the meanings at (a)–(f ) are possible, and only the
context can disambiguate them (Bruce : –). The options are summarized in a
tabular form.
. ritm fiñji tandhi-ak-ni-r-më-t-m Alamblak
insects  roast-get-go--.-sgf-pl
scope of
negation
roast get go
(a) ‘She did not roast (and) get the insects (and) go’ v v v
(b) ‘She took the insects unroasted’ v
(c) ‘She roasted the insects and went having left them’ v
(didn’t take them)
(d) ‘She roasted and got the insects but did not go’ v
(e) ‘She left the insects uncooked and went’ v v
(f) ‘She roasted the insects, didn’t take them and didn’t go’ v v
Components of multi-word serial verb constructions in Tariana must have the same
person marking. If a serial verb construction is negated, the negator (prefix ma- and
suffix -kade) goes onto the first component (V), negating the whole construction.
The subsequent components keep the person-number-gender cross-referencing. An
example of a negated causative serial verb construction is at ., from a story about a
spirit who ordered the man not to consume a drink which had been laced with
poison. The sentence can be translated as either (a) ‘He ordered him not to drink’, or
(b) ‘He didn’t order him to drink’. Only the first translation, (a), makes sense within
the context of the story.
. pa-ira-nipe-nuku ma-ra-kade-pidana Tariana
-drink--../ -order--..
dira di-na
masc.sg+drink masc-
(a) ‘He (the spirit) ordered him (the man) not to drink the drink’ (order drink)
(b) ‘He (the spirit) didn’t order him to drink the drink’9
Another random document with
no related content on Scribd:
 This eBook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this eBook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

1.E.2. If an individual Project Gutenberg™ electronic work is derived

from texts not protected by U.S. copyright law (does not contain a
notice indicating that it is posted with permission of the copyright
holder), the work can be copied and distributed to anyone in the
United States without paying any fees or charges. If you are
redistributing or providing access to a work with the phrase “Project
Gutenberg” associated with or appearing on the work, you must
comply either with the requirements of paragraphs 1.E.1 through
1.E.7 or obtain permission for the use of the work and the Project
Gutenberg™ trademark as set forth in paragraphs 1.E.8 or 1.E.9.

1.E.3. If an individual Project Gutenberg™ electronic work is posted

with the permission of the copyright holder, your use and distribution
must comply with both paragraphs 1.E.1 through 1.E.7 and any
additional terms imposed by the copyright holder. Additional terms
will be linked to the Project Gutenberg™ License for all works posted
with the permission of the copyright holder found at the beginning of
this work.

1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files containing a
part of this work or any other work associated with Project
Gutenberg™.

1.E.5. Do not copy, display, perform, distribute or redistribute this

electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1 with
active links or immediate access to the full terms of the Project
Gutenberg™ License.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if you
provide access to or distribute copies of a Project Gutenberg™ work
in a format other than “Plain Vanilla ASCII” or other format used in
the official version posted on the official Project Gutenberg™ website
(www.gutenberg.org), you must, at no additional cost, fee or expense
to the user, provide a copy, a means of exporting a copy, or a means
of obtaining a copy upon request, of the work in its original “Plain
Vanilla ASCII” or other form. Any alternate format must include the
full Project Gutenberg™ License as specified in paragraph 1.E.1.

1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™ works
unless you comply with paragraph 1.E.8 or 1.E.9.

1.E.8. You may charge a reasonable fee for copies of or providing

access to or distributing Project Gutenberg™ electronic works
provided that:

• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”

• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt that
s/he does not agree to the terms of the full Project Gutenberg™
License. You must require such a user to return or destroy all
 copies of the works possessed in a physical medium and
discontinue all use of and all access to other copies of Project
Gutenberg™ works.

• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.

• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.

1.E.9. If you wish to charge a fee or distribute a Project Gutenberg™

electronic work or group of works on different terms than are set
forth in this agreement, you must obtain permission in writing from
the Project Gutenberg Literary Archive Foundation, the manager of
the Project Gutenberg™ trademark. Contact the Foundation as set
forth in Section 3 below.

1.F.

1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on, transcribe
and proofread works not protected by U.S. copyright law in creating
the Project Gutenberg™ collection. Despite these efforts, Project
Gutenberg™ electronic works, and the medium on which they may
be stored, may contain “Defects,” such as, but not limited to,
incomplete, inaccurate or corrupt data, transcription errors, a
copyright or other intellectual property infringement, a defective or
damaged disk or other medium, a computer virus, or computer
codes that damage or cannot be read by your equipment.

1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except

for the “Right of Replacement or Refund” described in paragraph
1.F.3, the Project Gutenberg Literary Archive Foundation, the owner
of the Project Gutenberg™ trademark, and any other party
distributing a Project Gutenberg™ electronic work under this
agreement, disclaim all liability to you for damages, costs and
expenses, including legal fees. YOU AGREE THAT YOU HAVE NO
REMEDIES FOR NEGLIGENCE, STRICT LIABILITY, BREACH OF
WARRANTY OR BREACH OF CONTRACT EXCEPT THOSE
PROVIDED IN PARAGRAPH 1.F.3. YOU AGREE THAT THE
FOUNDATION, THE TRADEMARK OWNER, AND ANY
DISTRIBUTOR UNDER THIS AGREEMENT WILL NOT BE LIABLE
TO YOU FOR ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL,
PUNITIVE OR INCIDENTAL DAMAGES EVEN IF YOU GIVE
NOTICE OF THE POSSIBILITY OF SUCH DAMAGE.

1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you

discover a defect in this electronic work within 90 days of receiving it,
you can receive a refund of the money (if any) you paid for it by
sending a written explanation to the person you received the work
from. If you received the work on a physical medium, you must
return the medium with your written explanation. The person or entity
that provided you with the defective work may elect to provide a
replacement copy in lieu of a refund. If you received the work
electronically, the person or entity providing it to you may choose to
give you a second opportunity to receive the work electronically in
lieu of a refund. If the second copy is also defective, you may
demand a refund in writing without further opportunities to fix the
problem.

1.F.4. Except for the limited right of replacement or refund set forth in
paragraph 1.F.3, this work is provided to you ‘AS-IS’, WITH NO
OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.

1.F.5. Some states do not allow disclaimers of certain implied

warranties or the exclusion or limitation of certain types of damages.
If any disclaimer or limitation set forth in this agreement violates the
law of the state applicable to this agreement, the agreement shall be
interpreted to make the maximum disclaimer or limitation permitted
by the applicable state law. The invalidity or unenforceability of any
provision of this agreement shall not void the remaining provisions.
1.F.6. INDEMNITY - You agree to indemnify and hold the
Foundation, the trademark owner, any agent or employee of the
Foundation, anyone providing copies of Project Gutenberg™
electronic works in accordance with this agreement, and any
volunteers associated with the production, promotion and distribution
of Project Gutenberg™ electronic works, harmless from all liability,
costs and expenses, including legal fees, that arise directly or
indirectly from any of the following which you do or cause to occur:
(a) distribution of this or any Project Gutenberg™ work, (b)
alteration, modification, or additions or deletions to any Project
Gutenberg™ work, and (c) any Defect you cause.

Section 2. Information about the Mission of

Project Gutenberg™
Project Gutenberg™ is synonymous with the free distribution of
electronic works in formats readable by the widest variety of
computers including obsolete, old, middle-aged and new computers.
It exists because of the efforts of hundreds of volunteers and
donations from people in all walks of life.

Volunteers and financial support to provide volunteers with the

assistance they need are critical to reaching Project Gutenberg™’s
goals and ensuring that the Project Gutenberg™ collection will
remain freely available for generations to come. In 2001, the Project
Gutenberg Literary Archive Foundation was created to provide a
secure and permanent future for Project Gutenberg™ and future
generations. To learn more about the Project Gutenberg Literary
Archive Foundation and how your efforts and donations can help,
see Sections 3 and 4 and the Foundation information page at
www.gutenberg.org.

Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-profit
501(c)(3) educational corporation organized under the laws of the
state of Mississippi and granted tax exempt status by the Internal
Revenue Service. The Foundation’s EIN or federal tax identification
number is 64-6221541. Contributions to the Project Gutenberg
Literary Archive Foundation are tax deductible to the full extent
permitted by U.S. federal laws and your state’s laws.

The Foundation’s business office is located at 809 North 1500 West,

Salt Lake City, UT 84116, (801) 596-1887. Email contact links and up
to date contact information can be found at the Foundation’s website
and official page at www.gutenberg.org/contact

Section 4. Information about Donations to

the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission of
increasing the number of public domain and licensed works that can
be freely distributed in machine-readable form accessible by the
widest array of equipment including outdated equipment. Many small
donations ($1 to $5,000) are particularly important to maintaining tax
exempt status with the IRS.

The Foundation is committed to complying with the laws regulating

charities and charitable donations in all 50 states of the United
States. Compliance requirements are not uniform and it takes a
considerable effort, much paperwork and many fees to meet and
keep up with these requirements. We do not solicit donations in
locations where we have not received written confirmation of
compliance. To SEND DONATIONS or determine the status of
compliance for any particular state visit www.gutenberg.org/donate.

While we cannot and do not solicit contributions from states where

we have not met the solicitation requirements, we know of no
prohibition against accepting unsolicited donations from donors in
such states who approach us with offers to donate.

International donations are gratefully accepted, but we cannot make

any statements concerning tax treatment of donations received from
outside the United States. U.S. laws alone swamp our small staff.

Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.

Section 5. General Information About Project

Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could be
freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose network of
volunteer support.

Project Gutenberg™ eBooks are often created from several printed

editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.

Most people start at our website which has the main PG search
facility: www.gutenberg.org.

This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.