Discourse Markers and Beyond Descriptive and Critical Perspectives On Discourse Pragmatic Devices Across Genres and Languages 1St Ed 2020 Edition Peter B Furko Full Chapter PDF

Discourse Markers and Beyond:
Descriptive and Critical Perspectives

on Discourse-Pragmatic Devices
across Genres and Languages 1st ed.
2020 Edition Péter B. Furkó
Visit to download the full and correct content document:
https://ebookmass.com/product/discourse-markers-and-beyond-descriptive-and-critic
al-perspectives-on-discourse-pragmatic-devices-across-genres-and-languages-1st-ed
-2020-edition-peter-b-furko/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
A Contrastive View of Discourse Markers: Discourse

Markers of Saying in English and French 1st ed. 2020
Edition Laure Lansari
https://ebookmass.com/product/a-contrastive-view-of-discourse-
markers-discourse-markers-of-saying-in-english-and-french-1st-
ed-2020-edition-laure-lansari/
Discourse Markers: An Enunciative Approach 1st Edition

Graham Ranger
https://ebookmass.com/product/discourse-markers-an-enunciative-
approach-1st-edition-graham-ranger/
Language, Vernacular Discourse and Nationalisms 1st ed.

Edition Finex Ndhlovu
https://ebookmass.com/product/language-vernacular-discourse-and-
nationalisms-1st-ed-edition-finex-ndhlovu/
Ethical Discourse in Finance: Interdisciplinary and

Diverse Perspectives Marizah Minhat
https://ebookmass.com/product/ethical-discourse-in-finance-
interdisciplinary-and-diverse-perspectives-marizah-minhat/
Therapy as Discourse: Practice and Research 1st ed.
Edition Olga Smoliak
https://ebookmass.com/product/therapy-as-discourse-practice-and-
research-1st-ed-edition-olga-smoliak/
Foreign Languages in Advertising: Linguistic and

Marketing Perspectives 1st ed. 2020 Edition Jos Hornikx
https://ebookmass.com/product/foreign-languages-in-advertising-
linguistic-and-marketing-perspectives-1st-ed-2020-edition-jos-
hornikx/
Expressions of War in Australia and the Pacific:

Language, Trauma, Memory, and Official Discourse 1st
ed. 2020 Edition Amanda Laugesen
https://ebookmass.com/product/expressions-of-war-in-australia-
and-the-pacific-language-trauma-memory-and-official-
discourse-1st-ed-2020-edition-amanda-laugesen/
The Politics of Climate Change Metaphors in the U.S.

Discourse: Conceptual Metaphor Theory and Analysis from
an Ecolinguistics and Critical Discourse Analysis
Perspective Othman Khalid Al-Shboul
https://ebookmass.com/product/the-politics-of-climate-change-
metaphors-in-the-u-s-discourse-conceptual-metaphor-theory-and-
analysis-from-an-ecolinguistics-and-critical-discourse-analysis-
perspective-othman-khalid-al-shboul/
Environmental Organizations and Reasoned Discourse

Richard M. Robinson
https://ebookmass.com/product/environmental-organizations-and-
reasoned-discourse-richard-m-robinson/
POSTDISCIPLINARY STUDIES IN DISCOURSE
SERIES EDITOR: JOHANNES ANGERMULLER
Discourse Markers
and Beyond
on Discourse-Pragmatic Devices across
Genres and Languages
Péter B. Furkó
Postdisciplinary Studies in Discourse
Series Editor
Johannes Angermuller
Centre for Applied Linguistics
University of Warwick
Coventry, UK
Postdisciplinary Studies in Discourse engages in the exchange between
discourse theory and analysis while putting emphasis on the intellectual
challenges in discourse research. Moving beyond disciplinary divisions
in today’s social sciences, the contributions deal with critical issues at
the intersections between language and society.
Edited by Johannes Angermuller together with members of
DiscourseNet, the series welcomes high-quality manuscripts in dis-
course research from all disciplinary and geographical backgrounds.
DiscourseNet is an international and interdisciplinary network of
researchers which is open to discourse analysts and theorists from all
backgrounds.
Editorial Board
Cristina Arancibia
Aurora Fragonara
Péter Furkó
Tian Hailong
Jens Maesse
Eduardo Chávez Herrera
Michael Kranert
Jan Krasni
María Laura Pardo
Yannik Porsché
Kaushalya Perera
Luciana Radut-Gaghi
Marco Antonio Ruiz
Jan Zienkowski
More information about this series at

http://www.palgrave.com/gp/series/14534
Péter B. Furkó
Discourse Markers
and Beyond
on Discourse-Pragmatic Devices across
Genres and Languages
Péter B. Furkó
Department of English Linguistics
Károli Gáspár University of the Reformed
Church in Hungary
Budapest, Hungary
Postdisciplinary Studies in Discourse

ISBN 978-3-030-37762-5 ISBN 978-3-030-37763-2 (eBook)
https://doi.org/10.1007/978-3-030-37763-2
© The Editor(s) (if applicable) and The Author(s) 2020

This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by
similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with
regard to jurisdictional claims in published maps and institutional affiliations.
Cover illustration: RooM the Agency/Alamy Stock Photo
This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Contents
1 Preliminary Issues: Category Membership, Methodology,

Alternative Perspectives on Discourse Markers 1
1 Theoretical and Empirical Approaches to Discourse Markers 1
2 Characteristics of Discourse Markers, Criteria for
Discourse Marker Status 6
2.1 Non-propositionality and Optionality 6
2.2 Context-Dependence 8
2.3 Multifunctionality 8
2.4 Weak Clause Association and Variable Scope 9
2.5 Procedural Meaning/Non-compositionality 12
2.6 High Frequency, Orality, Stigmatization 13
3 Automatic Semantic Annotation: Testing Its Methods
and Precision 15
4 Corpus and Methodology 16
5 Findings 18
6 Conclusions, Utility of USAS as a Heuristic Tool 27
7 Alternative Perspectives on Discourse-Pragmatic
Devices: Outline of the Volume 28
References 32
v
vi Contents
Part I Discourse Markers Across Genres
2 Discourse Markers in Natural Conversations,

Scripted Conversations and Political Interviews:
Core and Peripheral Uses 39
1 Introduction 39
1.1 Types of Reports: Direct Reports, Indirect
Reports and Voicing 40
1.2 Research Questions 42
2 Research Material 43
3 Automated Semantic Annotation and Keyness of
Reporting Verbs and Expressions 46
4 The Use of Discourse Markers in Reporting Based
on the Findings of Manual Annotation 50
5 Prototypical and Peripheral Uses of Discourse
Markers in Reporting Across Four Genres 52
5.1 Core and Peripheral Examples from
the MPI Sub-corpus 52
5.2 Core and Peripheral Examples from Scripted
Discourse 56
5.3 Core and Peripheral Examples from Celebrity
Interviews and Natural Conversations 57
6 Summary, Conclusions 59
References 61
3 Discourse Markers from a Critical Perspective:

Some Theoretical Issues 65
1 Introduction 65
1.1 Discourse Marker Research and Its Relevance
to (Critical) Discourse Theory 66
1.2 Perspectives on the Manipulative Potential
of Discourse Markers 67
2 Previous Micro-Analyses of Manipulative Intent
in Political Discourse 69
Contents vii
3 The Political News Interview as a Genre 71

4 Data and Methodology 72
5 Discourse Markers Marking Manipulative Intent
in Political Interviews 74
5.1 Evidential Markers: Suppression and Backgrounding 74
5.2 General Extenders: Playing Down the Importance
of Alternative Viewpoints 77
5.3 The Role of Quotation Markers
in Decontextualizing and Recontextualizing
Texts, Legitimizing Opinions and Polarizing
the Audience 79
5.4 Other Manifestations of Manipulation:
Conversationalization and the Exploitation
of Ambiguity 81
6 Conclusions, Directions for Further Research 83
References 85
4 Discourse Markers from a Critical Perspective:

A Case Study of Discourse Markers in Parliamentary
Speeches 91
1 Introduction: Populism and Populist Discursive
Strategies 91
2 Contextual Background 93
3 Research Questions, Corpus and Methodology 95
4 Characteristics of Parliamentary Speeches 96
5 Propositional Lexical Items and Indexicals Used
as Manifestations of Populist Discursive Strategies 98
6 Discourse Markers and Modal Adverbs Used as
Manifestations of Populist Discursive Strategies 102
References 114
viii Contents
Part II Discourse Markers Across Languages
5 The Use of Discourse Markers in Business English

Textbooks: Issues in L2 Communicative Competence
and Learners’ Input 119
1 Introduction 119
2 Components of Communicative Competence 120
3 The Role of Discourse Markers in Shaping Learners’
Communicative Competence 122
4 Characteristics of Discourse Markers as Sources
of Learners’ Difficulties 125
5 Mapping the Functional Spectrum of Discourse
Markers in a Corpus of Business English Textbooks 126
5.1 Research Process 126
5.2 The Functional Spectrum of Well in Naturally
Occurring Discourse 127
5.3 The Functional Spectrum of Of Course in
Naturally Occurring Discourse 129
5.4 The Functional Spectrum of Well and Of Course
in a Corpus of Business English Textbooks 130
6 Conclusion 136
References 138
6 Discourse Markers in Scripted Discourse I: Issues

of (Under)Specification in the Translation
of Reformulation Markers 141
1 Introduction 141
1.1 Contrastive Approaches to Discourse Markers 142
2 Preliminary Considerations 143
2.1 The Translation of Discourse Markers—Theory
and Practice 143
2.2 From Scripted Discourse to the Language of
Subtitles 146
3 Previous Accounts of Reformulation 148
3.1 Definitions and Lists of Reformulation Markers 148
Contents ix
3.2 The Functional Spectrum of I Mean 151

3.3 The Functional Spectrum of Actually 152
4 The Study 153
4.1 Data and Methodology 153
4.2 Findings 155
4.3 Discussion, Translation Strategies
and Implications 157
5 Conclusions, Limitations and Directions
for Further Research 159
References 160
7 Discourse Markers in Scripted Discourse II:

The Representation and Translation of Irish English
Stereotypes in Contemporary Cinematography 165
1 Introduction 165
1.1 The Concepts of “Stereotype” and “Stereotyping” 166
1.2 “Stereotype” and “Stereotyping” in Interactional
Sociolinguistics and Sociopragmatics 168
2 Examples of Discourse-Pragmatic Stereotyping 169
2.1 Request Strategies 169
2.2 Compliment Responses (CRs) 171
2.3 Responses to Thanks (RTs) 173
3 Stereotypes Incurred by the Use of Discourse Markers 174
4 Quantitative Perspectives on IrE Stereotypes 179
4.1 Keyness Analysis and Automated Semantic
Tagging of IrE Scripted Discourse
with Reference to AmE Scripted Dialogues 179
4.2 Keyness Analysis and Automated Semantic
Tagging of IrE Scripted Discourse
with Reference to BrE Scripted Dialogues 181
5 Conclusions, Fuzzy Boundaries Between
Pragmalinguistic and Sociopragmatic Features
as Well as Sociocultural Norms 183
References 184
x Contents
8 Discourse Markers and Their Translation in Literary

Discourse: A Case Study of Discourse-Pragmatic
Devices in The Hobbit 187
1 Introduction 187
1.1 Literary Pragmatics—Perspectives and Approaches 188
1.2 Perspectives on Tolkien’s Linguistic Beliefs
and Style 189
2 Metapragmatic Awareness and Metacommunicative
Reflections in Tolkien’s Novels 190
3 The Functional Spectrum of Discourse Markers
in The Hobbit 195
3.1 The Use of Evidential Markers 195
3.2 Reformulation Markers—Self-Correction
and Mock Technicality 198
3.3 “… the Gold and the Dragon, and All
That”—General Extenders in The Hobbit 200
4 Quantitative Perspectives on Tolkien’s Literary
Style and Authentication Strategies 202
5 Some Issues Regarding the Translation
of Authentication Strategies 206
5.1 Strategies and Creativity in the Translation
of Discourse Markers in The Hobbit 208
5.2 The Translation of Speech Community
Creating Devices in the LOR Trilogy 212
6 Conclusions, Tolkien’s Philosophy of Language
Reconsidered 214
References 215
9 The Use of Discourse Markers in Bible Translations 219

1 Introduction, the Contribution of Discourse Marker
Research to the Study of New Testament Sentence
Conjunctions 219
1.1 Categorization and Typologies 220
1.2 Optionality and Stigmatization 222
Contents xi
2 Greek Sentence Conjunctions in John and Galatians

and Their Corresponding Discourse Markers
in Various Bible Translations 225
2.1 The Use of καί in Narrative and Expository Bible
Texts 226
2.2 The Functional Spectrum of δέ 228
2.3 The Functional Spectrum of γάρ 231
2.4 ἀλλά as a Global Marker of Contrast 234
3 Quantitative Perspectives on the Formal and Functional
Equivalence of New Testament Discourse Markers 235
References 242
Concluding Remarks 247
Software and Online Tools 251
Data Availability 253
Appendix—List of Abbreviations 255
References 257
Index 283
List of Figures
Chapter 2
Fig. 1 Concordance plots of Q2.1 and Q2.2 tags across
the four sub-corpora (Source Adopted from Furkó et al.
[2019: 253]) 48
Fig. 2 Presence and absence of discourse markers in reporting
across the four corpora (Source Adopted from Furkó et al.
[2019: 253]) 51
Chapter 3
Fig. 1 Types of reporting across four genres: scripted discourse
(SD), mediatized political interviews (MPI), celebrity
interviews (CI) and natural conversation (NC)
(Source Adopted from Furkó et al. [2019: 263]) 81
Chapter 8
Fig. 1 Translations of well in two Hungarian editions of The Hobbit 210
Fig. 2 Translations of of course in two Hungarian editions
of The Hobbit 210
xiii
xiv List of Figures
Chapter 9
Fig. 1 Specification and underspecification in 150 verses of Galatians 237
Fig. 2 Specification and underspecification in 150 verses of John 237
List of Tables
Chapter 1
Table 1 Alternative terms for English discourse-pragmatic devices 3
Table 2 Individual authors’ focus on categorial properties discourse
markers display 5
Table 3 Summary of discourse marker and non-discourse
marker-related semantic tags assigned to the most frequent
discourse marker types in the MPI and CI sub-corpora 19
Table 4 Inter-annotator agreement between automated and manual
tagging of discourse marker/non-discourse marker tokens 21
Chapter 2
Table 1 Semantic fields in USAS 46
Table 2 Normalized frequencies of the USAS categories
relevant to reporting 47
xv
xvi List of Tables
Chapter 4
Table 1 Frequency and keyness of potentially populist uses
of lexical items and suffixes in speeches given by governing
and opposition parties in the period between 8 May 2018
and 18 September 2019 99
Table 2 Frequency and keyness of potentially populist uses
of lexical items and suffixes in speeches given
by governing and opposition parties surrounding
the immigration quota referendum of 2016 100
Table 3 Frequency and keyness of potentially manipulative
discourse markers in speeches given by members
of governing and opposition parties in the period
of 8 May 2018 and 18 September 2019 104
Table 4 Frequency and keyness of potentially manipulative
discourse markers in speeches given by members
of governing and opposition parties 105
Chapter 5
Table 1 The use and functional spectrum of well and of course
in BE textbooks adopted from Furkó and Mónos
(2013: 142–143) 132
Table 2 The use and functional spectrum of well and of course
in BE textbooks analysed in this chapter 134
Chapter 6
Table 1 RMs listed by author and year of publication 150
Chapter 7
Table 1 Keyness analysis of lexical items in the IEC and SD
sorted by test corpus and keyness 180
Table 2 Keyness analysis of lexical items in the IEC and BSD
sorted by test corpus and keyness 182
List of Tables xvii
Chapter 8
Table 1 Lexical items associated with authentication and
the pragmatic mode, sorted by test corpus and keyness 204
Table 2 Formal indicators of the use of the pragmatic mode
in the THC and the CSLC 205
Table 3 Frequency of USAS tags associated with textual,
interactional and subjectivity markers in the THC
and the CSLC sorted by log-likelihood 207
Table 4 Hungarian discourse markers associated with
spontaneous conversations in two Hungarian editions
of The Hobbit 211
Table 5 Pragmatic routines and their translation based
on Bayona (2003: 81) 213
Chapter 9
Table 1 Translation Equivalents of kαί (TEs in each row
are listed in order of frequency) 227
Table 2 Translation equivalents of δε (TT discourse markers
in each row are listed in order of frequency) 232
Table 3 Frequency of specification and underspecification
underspecification strategies in the various Bible
translations 236
Table 4 The frequency of and, but and for in the KJV and the ASV 238
Table 5 Frequency of USAS tags associated with textual,
interactional and subjectivity markers across the KJV
and the ASV of the New Testament sorted by log-likelihood 239
1
Preliminary Issues: Category Membership,
Methodology, Alternative Perspectives
on Discourse Markers
1 Theoretical and Empirical Approaches

to Discourse Markers
The present volume is informed by research in a sub-field of pragmatics,
discourse marker research, often considered a “growth industry” (Fraser
1999: 931) and, at the same time, a “testing ground” (Bordería 2008:
1354) for pragmatic theories.
Discourse markers comprise a functional class of linguistic items that
do not typically change the propositional meaning of an utterance but
are essential for the organization and structuring of discourse, for mark-
ing the speaker’s attitudes to the proposition being expressed as well as for
facilitating processes of pragmatic inferences. A variety of approaches and
definitions have been offered, each one informed by a particular theo-
retical framework (Conversation Analysis, Interactional Sociolinguistics,
Rhetorical Structure Theory, Relevance Theory, etc., for an overview cf.
Fischer 2006; Furkó 2007; Vaskó 2016; Dér 2017).
Despite the rapidly growing body of discourse marker research, experts
in the field observe over and over again that there are still a number of
fundamental questions that need to be answered (cf. e.g. Schourup 1999;
Fraser 1999; Dér 2010, 2017; Heine 2013; Vaskó 2016). Some of the
© The Author(s) 2020 1
P. B. Furkó, Discourse Markers and Beyond, Postdisciplinary Studies
in Discourse, https://doi.org/10.1007/978-3-030-37763-2_1
2 P. B. Furkó
issues include the lack of generally accepted terminology and classifica-

tions and uncertainty regarding essential formal, semantic and pragmatic
characteristics.
The resulting terminological turmoil is illustrated in Table 1, which
provides a summary of some of the terms in classical studies and more
recent analyses referring to the respective discourse uses of well, of course,
oh, etc.:
The issue of terminology, alternative terms and overlapping cate-
gories such as pragmatic markers, connectives, contextualization cues and
enunciative markers will be the focus of Chapter 3 of the volume, while a
number of monographs (cf. Furkó 2007; Beeching 2016; Dér 2017) also
provide detailed discussions of the theoretical assumptions behind the
use of competing terms. Furkó (2007), in particular, takes the pragmatic
marker-discourse marker dichotomy as its major theme. Dér (2017: 9)
observes that the term discourse marker appears to be the most inclusive
and frequently used in the English literature, while its Hungarian mirror
translation “diskurzusjelölő” also appears to be the most widespread in
the respective academic community (Dér 2017: 10). Accordingly, I will
be using this as an umbrella term in the present introduction and make
the distinction between textual (prototypically discourse) functions and
(inter)subjective (prototypically pragmatic) functions prominent in sub-
sequent chapters where relevant.
The present volume will provide a series of empirical case studies on
the use of discourse markers across genres and languages as well as critical
analyses of the manipulative potential of discourse markers. Therefore,
the present introduction will focus on the second problem area men-
tioned above, i.e. the issue of categorization and category membership
as well as the methods by which one can identify lexical items that are
discourse markers and distinguish them from non-discourse marker uses
of the source categories.
Describing the characteristics of the functional class of discourse mark-
ers and developing criteria for deciding for every given instance whether
or not it is a discourse marker have been major preoccupations in recent
discourse marker research. Authors usually provide exhaustive lists of the
formal, functional and stylistic features that are associated with discourse
Table 1 Alternative terms for English discourse-pragmatic devices
well of course oh ah now I mean but you know
Schourup (1985) DP – DP DP DP DP – DP
Schiffrin (1987) DM – DM – DM DM DM DM
Erman (1987) PE – – – – PE – PE
Fraser (1990) DM/pause DM int/pause marker pause marker DM ‘90-PM DM PM
and (1999) marker* ‘99-DM
James (1974) int – int int – – – –
Wierzbicka – – int int – – – –
(1991)
Hirschberg and cue – – – cue – cue –
Litman (1993)
Stenström DM/IS* – IS – DM/IS* IS – DM/IS*
(1994)
Holmes (1995) PP PP – – – PP – PP
Kroon (1995) DP – – – – DP DP DP
Nikula (1996) PFM PFM – – – PFM – PFM
Fuller (2003) DM DM DM – – DM – DM
Beeching (2016) PM booster – – – PM – PM
Crible (2017) DM – DM int DM DM DM DM
Legend
DP—discourse particle
DM—discourse marker
PE—pragmatic expression
PM—pragmatic marker
int—interjection
cue—cue word/cue phrase
1 Preliminary Issues: Category Membership, Methodology …
IS—interactional signal
PP—pragmatic particle
PFM—pragmatic force modifier
*—categorized according to the position/slot they take in the utterance
3
4 P. B. Furkó
markers as a functional class (cf. e.g. Schourup 1999; Fraser 1999; Beech-
ing 2016; Brinton 2017), but empirical studies rely on different subsets
of such criterial features when identifying particular instances of dis-
course markers in a given corpus (for a detailed discussion, cf. Crible
2017). Naturally, this makes it difficult to compare the results of empir-
ical research even if similar datasets are involved. Table 2 illustrates this
problem.
An even more challenging task is to develop annotation software that
can automatically identify discourse markers in oral discourse and fil-
ter out non-discourse marker tokens of lexical items that are frequently
used as discourse marker types (e.g. adverbial uses of well or now, prepo-
sitional uses of like, etc.). Moreover, to date, few attempts have been
made to use automated means of identification involving semantic crite-
ria and semantic fields, since one of the very criterial features of discourse
markers is their semantic underspecification (cf. Crible et al. 2019), which
is a result of the diachronic process of semantic bleaching (cf. Brinton
2017: 31).
Accordingly, the present introduction will explore the utility of
using an automated semantic tagging software, USAS as a pre-
annotation tool for the identification of oral discourse markers, includ-
ing (inter)subjective as well as textual markers. After an overview of the
formal and functional features that can be used for manual annotation,
and after comparing the results of manual and automatic annotation of
selected discourse markers, I will argue that despite the semantic under-
specification of most discourse markers, automatic semantic annotation
(ASA) can be an effective tool for the disambiguation between discourse
marker and non-discourse marker uses with regard to certain items, but
needs to be complemented by extensive manual error correction and
filtering.
Table 2 Individual authors’ focus on categorial properties discourse markers display
seq. context oral synt. proced poly-func. attitude scope non-prop. inv.
Schiffrin (1987) x x x x (x)
Fraser (1990, 1999) x x x x x
Redeker (1990, x x (x)
1991)
Stenström (1994) x (x)
Kroon (1995) x x x
Knott and Sanders x
(1998)
Andersen (1998) x x x
Hansen (1998) x (x) (x) x x
Risselada and x x
Spooren (1998)
Romaine and Lange x (x) (x)
(1998)
Blakemore (1987, x
2002)
González (2004) x x x
Crible (2017) x x x
Legend
seq.—sequentiality-coherence-connectivity
context—context-dependence—context-coordination
oral.—orality
synt.—syntactic criteria (diversity, non-integration)
proced.—procedural meaning
poly-funct.—poly-functionality
attitude—marking attitudes
scope—variable scope, functional scope
non-prop.—non-propositional content
inv.—invariable form
5
6 P. B. Furkó
2 Characteristics of Discourse Markers,

Criteria for Discourse Marker Status
2.1 Non-propositionality and Optionality
Many scholars (cf. Schourup 1999) consider non-propositionality (non-

truth-conditionality) as a sine qua non for discourse marker status, yet
others include propositional items such as then and after that. While it is
generally agreed that certain discourse markers (e.g. well, however, etc.)
contribute nothing to the truth-conditions of the proposition expressed
by an utterance, the non-truth-conditionality of others (frankly, I think)
have generated a great deal of controversy (cf. Infantidou-Trouki 1992;
Brinton 2017: 127ff ).
Blakemore (1987: 106) argues that a distinction has to be made
between truth-conditional and non-truth-conditional meaning, on the
one hand, and conceptual vs. procedural meaning, on the other. Thus,
many of the controversies stem from the fact that certain scholars confuse
the two distinctions and use them interchangeably. Schourup (1999), for
example, uses the compositionality test to argue in favour of the truth-
conditionality of in addition:
(1a) Owens is a respected drama critic. I tell you in addition that she
has written …
(1b) Owens is a respected drama critic. In addition, she has written …
While in addition is indeed truth-conditional, the above test would pre-

dict that frankly is also truth-conditional. However, as Blakemore (2002)
would argue, discourse marker uses of frankly are non-truth-conditional,
but conceptual. It is, therefore, important to point out that the com-
positionality test will be a useful tool in deciding whether individual
discourse markers have conceptual or procedural meaning, the truth-
functionality of discourse markers is tested more efficiently in terms of
whether they retain their original meaning when embedded in if-clauses
or under the scope of factive connectives such as because:
1 Preliminary Issues: Category Membership, Methodology … 7
(2a) Allegedly / Obviously / Frankly, the cook has poisoned the soup.
(2b) If the cook has allegedly / ?obviously / *frankly poisoned the soup,
we can eat the meal without worrying.
(2c) We shouldn’t eat the soup, because the cook has
allegedly/?obviously/*frankly poisoned it.
The uncertainty with regard to whether or not obviously retains its orig-
inal meaning in (2c) suggests to many that the truth-functionality–non-
truth-functionality distinction should be viewed as a continuum, rather
than a dichotomy, which is consistent with the finding in grammatical-
ization theory that due to the diachronic grammaticalization processes
that are synchronically manifested in the use of discourse markers, there
is a fuzzy boundary between uses that are non-truth-conditional and
(omissible) and those that are not (for a detailed discussion, cf. Andersen
2001; Blakemore 2002; Dér 2017).
Optionality as a distinguishing feature is in many respects derivative of
the previously discussed criterion of non-propositionality, and discourse
markers are considered optional from the perspective of sentence mean-
ing because their absence does not change the conditions under which
the sentence is true.
There are, however, two further senses in which discourse markers are
claimed to be optional. Firstly, they may be seen as syntactically optional
in the sense that removal of a discourse marker does not alter the gram-
maticality of its host sentence. Secondly, they are optional in the sense
that if a discourse marker is omitted, the relationship it signals is still
available to the hearer, though no longer explicitly cued (cf. Schourup
1999: 231).
The above statement does not entail that discourse markers are useless;
rather, it reflects the view according to which discourse markers guide
the hearer towards a particular interpretation of the connection between
a sequence of utterances and at the same time rule out unintended inter-
pretations.
8 P. B. Furkó
2.2 Context-Dependence
Discourse markers’ extreme context-dependence is frequently identified

with their inherent indexicality. Aijmer, for example, considers indexi-
cality as the most important property of discourse markers, a property
whereby discourse markers are linked to attitudes, evaluation, types of
speakers and other features of the communicative situation (cf. Aijmer
2002: 5). In this respect, discourse markers can be compared to deic-
tics, i.e. another borderline phenomenon can be observed if we look at
some of the definitions of deictic expressions, which often overlap with
those of discourse markers. Both categories are usually defined in terms
of context-dependence, i.e. in terms of having meaning only by virtue
of an indexical connection to some aspect of the speech event (cf. e.g.
Sidnell 1998). Levinson (2004), in fact, considers discourse markers as
discourse deictics, other subgroups including spatial, temporal and social
deictics.
Similarities between indexicals and discourse markers are also recog-
nized by proponents of Relevance Theory. Carston, for example, notes
that the two seemingly disparate phenomena are brought together by
the fact that both encode a procedure rather than a concept, and both
play a role in guiding the hearer in the pragmatic inferential phase of
understanding an utterance (Carston 1998: 24). The difference between
the two sets of phenomena, according to Carston, is that indexicals con-
strain the inferential construction of explicatures and discourse markers
(discourse connectives in RT terms) constrain the derivation of implica-
tures (in other words, intended contextual assumptions and contextual
effects).
2.3 Multifunctionality
In addition to playing a role in pragmatic inferencing, individual dis-

course markers are also associated with a plethora of functions includ-
ing hedging and politeness functions. What is more, they can also be
salient in conversational exchanges as openers, turn-taking devices, hes-
itational devices, backchannels, markers of topic shift and of receipt of
information, and so on (cf. e.g. Beeching 2016: 4ff ). Discourse mark-

ers are inherently multifunctional and ambiguous, since there is a lot
of interpersonal and discourse burden on their signalling capacity. Dis-
course markers signal interpersonal and discourse functions simultane-
ously; thus, they are ambiguous between the two levels, and on the other
hand, they are vague with regard to signalling particular relations on a
given level as well (ibid.).
The multifunctionality of discourse markers also brings up the ques-
tion of whether different uses of a given marker are to be considered
incidental and unrelated (maximalist approach) or motivated and related
(minimalist approach) and whether there is an invariant “core mean-
ing” of discourse markers that is context-independent and preserves some
component of the lexeme’s original semantic meaning. Since the focus of
the present introduction is on differentiating between discourse marker
and non-discourse marker uses of a given item, further discussion will
not ensue on the multifunctionality of discourse marker uses. The the-
oretical issue will be picked up in Sect. 1.1 of Chapter 9 with regard to
connectives, often considered the most underspecified subgroup of dis-
course markers (cf. e.g. Crible 2018: 211). Multifunctionality will also be
an important consideration from an empirical perspective when develop-
ing coding schemes for manual annotation, as we will see in Sect. 4 of
Chapter 2 as well as most of the chapters in the volume.
2.4 Weak Clause Association and Variable Scope
It is frequently observed in the literature that discourse markers usually

occur either outside the syntactic structure or loosely attached to it (cf.
e.g. Crible 2017: 111). Quirk et al. classify many linguistic items that are
elsewhere included among discourse markers as conjuncts (e.g. nonethe-
less) which are considered to be clause elements but to have a detached
role relative to other, more closely interrelated clause elements such as
subject, complement and object:
10 P. B. Furkó
Conjuncts are more like disjuncts than adjuncts in having a relatively

detached and ‘superordinate’ role as compared with other clause elements
(Quirk et al. 1985: 631).
In addition, some of the items that Quirk et al. refer to as “disjuncts”

(e.g. obviously, sentence initial surprisingly and frankly) also display a
whole range of properties associated with the functional class of discourse
markers.
It is important to note that the property of weak clause association is
relative to elements external to the discourse marker’s lexical form, since
several discourse markers clearly have their own internal syntactic struc-
ture (e.g. on the other hand ) and others (e.g. y’know, I mean) constitute
clauses from a syntactic point of view despite the fact that they are no
longer considered to be compositional, but procedural (cf. e.g. Furkó
2014).
Weak clause association is frequently discussed in relation to phono-
logical independence: discourse markers often constitute independent
tone units or are set off from the main clause by “comma intonation”
(cf. Hansen 1997: 156).
Adding weak clause association and a corresponding lack of intonational
integration to our list of criteria could also be justified from the perspec-
tive of grammaticalization theory. An important clause of the definition
of grammaticalization states that it takes place in special morpho-syntactic
environments. In the case of discourse markers, this environment can
be associated with sentence-initial position, and hence, many scholars
regard quasi-initiality as yet another distinguishing feature of discourse
markers (cf. e.g. Schourup 1999). However, Dér (2017: 15) questions
the validity of these features on grounds of empirical evidence and distin-
guishes between different degrees of initiality. Moreover, once discourse
markers enter an advanced stage of grammaticalization, they become syn-
tactically independent and can appear at various parts of the sentence,
with an accompanying “comma intonation”, resulting in fuzzy bound-
aries and borderline cases in terms of syntactic dependence /indepen-
dence of individual tokens.
Discourse markers’ position in an utterance can also influence their
scope, which is variable, as is illustrated by (3a) and (3b):
(3a) Interviewer: I know how close you are to your mom. How old is
she?
Interviewee: Well, she probably doesn’t want me to say…
(3b) You’re not going to have quality if you can’t sleep and you itch
and you bitch and you weep and you cry and you bloat and you can’t
remember anything and you don’t have a, well, sex drive. (examples
taken from Furkó 2014)
As the examples above show, the size of the linguistic unit well can take
in its scope ranges from a whole sentence to a single word. Waltereit
(2006) observes that this variability is a remarkable property, but it is
not an exclusive feature of discourse markers, since conjunctions as a
word-class (and even some individual conjunctions as a lexical item) can
also have variable scope, giving the following sentences as examples:
(4a) Ed and Doris loved each other.

(4b) Ed worked at the barber’s, and Doris worked in a department
store.
In (4a), and has scope over two NPs, and in (4b), it has scope over two
clauses. However, the difference between and used as a conjunction and
its discourse marker use lies in the fact that the scope of the conjunction
and can always be determined in grammatical terms. It could be defined
as ranging over two constituents of the same type adjacent to and, which,
in turn, make up a constituent of again the same type. The scope of dis-
course markers, in contrast, cannot be determined in grammatical terms,
as is clear from (5) below:
(5) My husband got a notice t’go into the service

and we moved it up.
And my father died the week … after we got married.
And I just felt, that move was meant to be. (Schiffrin 1987: 53,
emphasis in the original)
Schiffrin (1987) concludes that and has “freedom of scope”, rather

than “variable scope”, since “we can no more use and to identify the
12 P. B. Furkó
interactional unit that is being continued than we can use and to identify
the idea that is being coordinated” (Schiffrin 1987:150).
Traugott (1995) relates the feature of variable scope to grammaticaliza-
tion and argues that in addition to nominal clines (nominal adposition >
case) and verbal clines (main verb > tense, aspect, mood marker), which
are “staples of grammaticalization theory”, a further cline: Clause inter-
nal Adverbial > Sentence Adverbial > Discourse Particle should be added
to the inventory (Traugott 1995: 1). According to Traugott, this cline
involves increased syntactic freedom and scope.
Brinton (2017: 24) further refines Traugott’s (1995) clines and adds
scope within the proposition > scope over the proposition > scope over dis-
course as a separate cline in the evolution from propositional to textual
and interpersonal meaning.
2.5 Procedural Meaning/Non-compositionality
Although most scholars treat non-compositionality as a property of dis-

course markers per se (for a detailed account, cf. Brinton 2017), Blake-
more (2002) associates discourse markers with procedural meaning and
uses non-compositionality as a test to decide whether individual items
are conceptual or procedural.
Blakemore also claims that if discourse markers are synonymous with
their non-discourse marker counterparts, they encode conceptual mean-
ing. Thus, seriously and in other words in (6a) and (7a) encode a concept
parallel to (6b) and (7b), respectively. On the other hand, well (as in 8a)
encodes a procedure, since it is not synonymous with well in (8b):
(6a) Seriously, you will have to leave.

(6b)He looked at me very seriously.
(7a)In other words, you’re banned.
(7b)She asked me to try and put it in other words.
(8a)A: What time should we leave?
B: Well, the train leaves at 11.23.
(8b)You haven’t ironed this very well.
A second test Blakemore uses is to see if a given item can combine

with linguistic items encoding conceptual meaning to produce complex
expressions.
As far as the question of synonymity is concerned, it is important to
note that the fact that, on the basis of native intuitions, no correspon-
dence can be found between the adverbial well and its discourse marker
counterpart, does not mean that such a relationship is absent (cf. e.g.
Furkó 2013). Native intuitions, naturally, disregard diachronic aspects of
individual lexical and grammatical items, and it is exactly these aspects
that account for the fuzziness of the category of discourse markers and
the resulting borderline cases.
2.6 High Frequency, Orality, Stigmatization
Last but not least, some of the stylistic features core members of the
functional class of discourse markers display need to be considered.
While semantic-functional properties are more important in determin-
ing class membership than formal and stylistic ones, stylistic criteria can
also be helpful in determining discourse marker status and differentiating
between discourse marker and non-discourse marker tokens.
It is important to note that high frequency of use is the backbone of
various processes of grammaticalization as well as pragmaticalization (cf.
e.g. Furkó 2014; Dér 2017). In other words, the more frequently an item
is used, the more likely it is that its formal-functional properties are going
to change, and once it has entered the process of grammaticalization, the
faster it is going to go through the substages of that process.
A number of studies on discourse markers observe that the frequency
of discourse markers can be primarily observed in speech (e.g. Beeching
2016); what is more, one of the most salient features of oral style is the
use of items such as well, right, ok and you know. For example, in their
classical study, Brown and Yule (1983: 17) label well, erm, I think, you
know, if you see what I mean, I mean, of course “prefabricated fillers”, when
drawing up a list of contrasting characteristics of spoken and written lan-
guage. They also point out that these items’ overuse is often stigmatized
by prescriptivists (ibid.).
14 P. B. Furkó
However, it is easy to illustrate the meaningfulness and distinctive (as

opposed to random) use of even the two most used discourse markers,
you know and I mean. As Fox Tree and Schrock (2002: 731) illustrate, it
matters where you know or I mean appears in an utterance and they are
not interchangeable:
(9a) Original: me and the Edinburgh girl got together after dinner
late in the evening and decided they’d really got us along to make
it look right, you know they had after all had candidates from other
universities.
Alternative: me and the Edinburgh girl got together after dinner you
know late in the evening and decided they’d really got us along to make
it look right, they had after all had candidates from other universities.
(9b) Original: but I don’t think it’s feasible. I mean I know this is the
first time I’ve done it, and I’m not in a main line paper, but I’m sure
it’ll take me all my time to do it in three weeks.
Alternative: but I don’t think it’s feasible. I know I mean this is the
first time I’ve done it, and I’m not in a main line paper, but I’m sure
it’ll take me all my time to do it in three weeks. (example taken from
Fox Tree and Schrock 2002: 731)
In (9a) Original, you know comments on what is meant by “look right”,

whereas in (9a) Alternative it comments on what “after dinner” means (in
other words, they differ in what they take within their scope, see Sect. 2.5
above). In (9b) Original, I mean comments on why the speaker says “I
don’t think it’s feasible”, without overwriting the statement, but in (9b)
Alternative, I mean comments on “I know”, retrospectively treating it as
a false start.
Moreover, as both manual and automatic annotation will illustrate,
there is no principled basis on which one could exclude from the func-
tional class of discourse markers connectives such as however, after all,
consequently and a range of other items characteristic of formal style,
some of which (e.g. besides, however, moreover ) are in fact included in
Brown and Yule’s (ibid.) list of the characteristics of written language.
The issue of overuse and stigmatization will be taken up in Sect. 3 of
Chapter 7.
3 Automatic Semantic Annotation:

Testing Its Methods and Precision
After the above discussion of the categorial features, discourse mark-
ers are usually associated with I cannot but agree with Crible (2017),
who argues that “any categorical definition is only useful insofar as it is
endorsed by an empirical model of identification and annotation” (2017:
99). She also provides an overview of most of the above features and
comes to the conclusion that they can be reduced to three criteria that are
helpful when identifying discourse marker types (1) and setting apart dis-
course marker tokens from non-discourse marker tokens (2&3): (1) mul-
tifunctionality, (2) syntactic (non-)integration and (3) functional scope,
respectively (Crible 2017: 105).
As for the issue of automatic annotation, there are a variety of comput-
erized semantic tagging (CST) systems, including artificial intelligence-
based, knowledge-based, corpus-based and semantic taxonomy-based
systems (for an overview, cf. e.g. Prentice 2010). By way of introduc-
ing the methodology used in the following chapters, the present analysis
draws on the results gained from the UCREL Semantic Analysis System
(USAS), which has the major advantage of combining these approaches.
Furthermore, USAS groups lexical items in terms of a taxonomy of
semantic fields and assigns semantic categories to all words, including
grammatical and other procedural (non-propositional) items, which is
relevant for the present dissertation in view of the fact that the lexical
items under scrutiny are highly procedural and semantically bleached (cf.
Sect. 2 above).
USAS system uses an automatic coding scheme of 21 semantic fields
(for details, cf. Table 1 in this Chapter), subdivided into 232 sub-
categories. For reasons of brevity, only the tags that have been asso-
ciated with the discourse marker types under analysis will be dis-
cussed in the present section, and the complete coding scheme can be
found at http://ucrel.lancs.ac.uk/usas/. USAS uses disambiguation meth-
ods including part-of-speech tagging, general likelihood ranking, multi-
word-expression extraction, domain of discourse identification and con-
textual rules (for a detailed discussion, cf. Rayson et al. 2004). Previous
16 P. B. Furkó
evaluations of the accuracy of the system reported a precision value of

91% (ibid.), i.e. a 9% margin of error applying to lexical items across
the board (including propositional and non-propositional items).
The research questions in the introductory study were as follows:
1. Are the disambiguation methods USAS uses sufficient for filtering out
non-discourse marker tokens of the most frequent discourse marker
types?
2. Does the margin of error reported to apply in general apply to the
identification of discourse markers as well?
3. Are individual discourse markers identified/tagged with a similar mar-
gin of error?
4. If individual discourse markers are tagged with varying precisions by
USAS, what formal-functional properties of the relevant discourse
markers might explain the differences?
4 Corpus and Methodology

In the course of the research, two sub-corpora of the same size (100,000
words each) were used:
• a corpus of the official transcripts of 39 confrontational type of medi-

atized political interviews (henceforth MPI sub-corpus) downsampled
from BBC’s Hard Talk and Newsnight (available at http://bbc.co.uk),
the transcripts comprise a total of 99,225 words ± 2%, allowing for
technical/transcript-specific information such as the indication of par-
ticipants’ names;
• a corpus of the official transcripts 50 celebrity interviews (henceforth
CI sub-corpus) downsampled from CNN’s Larry King Live (available
at http://www.cnn.com), each interview lasts approximately 50 min-
utes, the total word count (100,436 ± 2%) of the CNN sub-corpus
thus makes it comparable to the BBC sub-corpus.
To increase comparability, the downsampling technique (cf.

Khosravinik 2010) involved selecting the same time periods, similar
topics and themes of the interviews. Further genre-specific details about
the two sub-corpora will be provided in this Chapter of the volume. In
the present section, the results of automatic tagging will be compared
to findings based on manual annotation with respect to the D-values as
well as the functional distribution of individual lexical items (see Sect. 5
below).
The research process has been as follows: in order to identify and com-
pare the USAS tags of oral discourse markers in the two sub-corpora, the
semantic tags assigned to frequent discourse markers (e.g. I mean, you
know, in other words, so, well ) were considered, and then these semantic
tags were used to identify further types and tokens relevant to discourse
marking. It was found that 95.1% of the instances of discourse markers
trawled from the two sub-corpora through this method are either tagged
with Z4, described in the USAS manual as the “discourse bin” (including
items such as oh, I mean, you know, basically, obviously, right, yeah, yes) or
with A5.x, described as “evaluative terms depicting quality” (including
discourse markers such as well, OK, okay, good, right, alright ). The fre-
quency of the relevant tags across the two sub-corpora was compared,
as well as the ratio between discourse marker-relevant tags (i.e. Z4 and
A5.x) and non-discourse marker-relevant tags (e.g. B2, I1.1, T1.3, etc.,
see below for details).
In the second stage, a representative sample of 400 tokens in the MPI
sub-corpus was manually annotated using a numeric code of 1 for dis-
course marker and 2 for non-discourse marker tokens with a view to
comparing the results of automatic and manual tagging. When deciding
if an individual token is a discourse marker or not, Crible’s (2017) cri-
terial features (2017: 99, see also Sect. 2) were applied by a single expert
annotator. The tokens that were selected for the sample were weighted for
their frequency in the corpus, while discourse marker and non-discourse
marker tokens were included in equal proportions. For example, the 429
tokens of well comprise 19.6% of all automatically tagged items; thus,
78 tokens (39 A5.1-tagged and 39 non-A5.1 tagged by USAS) were
included in the sample.
18 P. B. Furkó
5 Findings
Table 3 summarizes the raw frequency of the relevant lexical items’ dis-
course marker and non-discourse marker-related USAS tags. Since both
sub-corpora were compiled in a way that they are of the same size of
100,000 words, the raw frequencies can also be compared as normalized
frequencies.
As a first step, the ratio of discourse marker and non-discourse marker
tokens of individual items was compared with the results of previous
research in the course of which discourse markers in the same sub-
corpora were manually annotated (cf. Furkó and Abuczki 2014 and
Sect. 3 in this Chapter). In order to gauge the categorial multifunction-
ality of discourse markers, the measure of D-function ratio or D-value
(a term proposed by Stenström 1990) was used. An individual item’s
D-value is calculated as a quotient of the number of tokens that ful-
fil discourse-pragmatic functions and the total number of occurrences
in a given corpus. The D-value of oh, for example, is 1 (100%) in the
London-Lund Corpus, since it is used exclusively as a discourse marker,
whereas well showed a D-value of 0.86 (86%), as 14% of its tokens serve
non-discourse marker (adverbial, nominal, etc.) functions (ibid.).
If we calculate the D-values of individual discourse markers based on
the above values and compare them to the findings of previous research,
we see that the results of automatic annotation and manual annotation
converge to a great extent. Mean, for example, has a D-value of 0.808 in
the MPI corpus based on automatic annotation (calculated as the num-
ber of Z4 tags divided by all tokens of mean, i.e. 141), while manual
annotation yielded a D-value of 0.797 (cf. Furkó and Abuczki 2014:
50). Similarly, manual annotation yielded a D-value of 0.82 for well in
the MPI corpus (Furkó and Abuczki 2014: 54), while Table 1 yields a
D-value of 0.839 for this lexical item (360 Z4 tags divided by the total
number of tokens, i.e. 429).
The table also correctly predicts that most of the lexical items under
scrutiny have higher D-values in the CI sub-corpus than in the MPI
sub-corpus, which is explained by the fact that there is a higher degree
of conversationalization in celebrity interviews, i.e. they are more similar
to spontaneous, informal, face-to-face conversations (cf. this chapter and
Table 3 Summary of discourse marker and non-discourse marker-related semantic tags assigned to the most frequent
discourse marker types in the MPI and CI sub-corpora
frequency of frequency of frequency of frequency of
DM-related tag in the DM-related tag in the non-DM-related tag non-DM-related tag
lexical item MPI CI in the MPI in the CI
well (429) 360xA5.1 312xA5.1 14xI1.1, 55xN5 1xA7, 2xB2, 24xN5
sort (38) 14xZ4 25xZ4 21xA4.1, 3xA1.1.1 10xA4.1
now (299) 4xZ4 1xZ4 288xT1.1.2, 7xZ5 229xT1.1.2, 6xZ5
(you) know (346) 205xZ4 455xZ4 140xX2.2, 1xZ6 307xX2.2
like (97) 6xZ4 17xZ4 51xZ5, 40xE2+ 238xZ5, 139xE2+
(I) mean (141) 114xZ4 201xZ4 27xQ1.1 30xQ1.1, 5xS2.2.2
(in other) words (11) 4xZ4 13xZ4 7xQ.3 7xQ.3
actually (165) 165xA5.4 72xA5.4 0 0
(I) think (549) 126xZ4 121xZ4 423xX2.1 319xX2.1
right (114) 55xZ4, 53xA5.3 211xZ4, 98xA5.3 6xT1.1.2 12xN3.8, 16xS7.4,
15xT1.1.2
19
20 P. B. Furkó
Chapter 2). For example, the D-value of well is 0.92, and the D-value
of mean is 0.851 in the CI sub-corpus based on automatic annotation
(312 A5.1 tags divided by a total of 339 tokens, 201 Z4 tags divided by
a total of 236 tokens, respectively).
In the second stage of the research, a representative sample of tokens in
the MPI was manually annotated using numeric 1 for discourse marker
tokens and 2 for non-discourse marker uses. With a view to comparing
the results of automatic and manual annotation, all discourse marker-
related tags (Z4 and A5.x) yielded by USAS were re-coded as numeric
1, while non-discourse marker tags (B2, I1.1, T1.3, etc.) were re-coded
as 2. Consequently, the extracted list of the corresponding manual and
automated tags was entered into a reliability calculator (Freelon’s ReCal
2 for 2 coders) in order to calculate inter-annotator agreement statistics.
Table 4 shows the result.
Although the above inter-coder agreement values appear high (cf.
Spooren and Degand 2010), it is important to note that there is a
great degree of variation in the precision with which individual dis-
course markers are tagged by USAS. On the one hand, there are dis-
course markers such as I mean and you know whose discourse marker and
non-discourse marker uses are disambiguated with surprising precision
(resulting in a kappa score of <.98, i.e. close to perfect inter-coder agree-
ment between USAS and the human annotator), cf. (10ab) and (11ab)
below:
(10a) I_Z4[i1.2.1 mean_Z4[i1.2.2 the_Z5 long-term_T1.3+

plans_X7 + or_Z5 Britain_Z2
are_A3+ for_Z5 a_Z5 second_N4 West_M7[i3.2.1 Coast_M7[i3.2.2
mainline_M3
railway_M3c ._PUNC
(10b) Well_A5.1+ ,_PUNC yes_Z4,_PUNC I _Z8mf do_Z5
mean_Q1.1 that_Z8 ._PUNC
(11a) Getting_A9+ crime_G2.1- down_Z5 below_Z5 what_Z8 it_Z8
used_A1.5.1,_PUNC
you_Z4[i1.2.1 know_Z4[i1.2.2,_PUNC otherwise_A6.1- would_A7+
be_A3+ _PUNC
(11b) You_Z8mf know_X2.2+ the_Z5 answer_Q2.2 to_Z5 that_Z5
question_Q2.2 ._PUNC
Table 4 Inter-annotator agreement between automated and manual tagging of discourse marker/non-discourse marker
tokens
Percent N
agreement Scott’s Pi Cohen’s Kappa N agreements disagreements N cases N decisions
Variable 92.75 0.85 0.85 371 29 400 800
(DM/non-DM)
21
22 P. B. Furkó
I mean in (10a) qualifies as a discourse marker on grounds of its non-

propositionality and syntactic optionality (cf. Sect. 2.1 above). In (10b),
on the other hand, I mean is not a discourse marker, as it receives an
object complement (that ) and the verbal status of mean is also under-
lined by do-support, which is used for emphasis. Similarly, (10a) contains
a discourse marker use of you know based on the criteria of syntactic inte-
gration and variable scope/discourse function. It is interesting to consider
that you know is correctly tagged despite the fact that the host utterance
does not strictly speak comprise a grammatical sentence, i.e. it does not
have a straightforward syntactic structure that can be parsed and PoS-
tagged in its original formulation. The utterance contains a false start
(below what it used ), a reformulation marker (the discourse marker you
know) and a reformulated segment (otherwise would be). The intended
utterance Getting crime down below what it otherwise would be is only
recoverable if the word used is not considered in the course of syntactic
and semantic parsing. You know in You know the answer to that question
is correctly tagged as a syntactically integrated, i.e. non-discourse marker
token.
The disambiguating precision with respect to you know and I mean
is probably due to two of the disambiguation methods USAS applies:
firstly, its multi-word-expression extraction algorithm and its core com-
ponent of MWE lexicon (cf. Rayson et al. 2004) and secondly, the fact
that POS tagging enables the parser to differentiate between syntactically
integrated tokens that are monotransitive (and are thus followed by their
nominal or clausal complements) and syntactically non-integrated ones
that are marked by the absence of complements.
However, even in the case of multi-word units, we can observe a
bias towards discourse marker-related tags, i.e. the discourse marker use
is overgeneralized and is applied to occurrences of I mean, I say and
you know that are syntactically integrated and have the valency pat-
tern/complement options of other monotransitive verbs, cf. (12) and
(13):
(12) I_Z8mf want_X7+ to_Z5 assure_A7+ you_Z8mf that_Z8

I _Z4[i1.2.1 mean_Z4[i1.2.2
what_Z8 I _Z4[i2.2.1 say _Z4[i2.2.2 when_Z5 I_Z8mf tell_Q2.2

you_Z8mf I_Z8mf
appreciate_E2+ your_Z8 contributions_A9- ._PUNC
(13) As_Z5 you_Z4[i1.2.1 know_Z4[i1.2.2,_PUNC the_Z5 Govern-
ment_G1.1c says_Q2.1
it_Z8 ’s_A3+ too_N5.2+ early_T4+ to_Z5 tell_Q2.2 about_Z5
that_Z8 ._PUNC
In (13), you know is not a discourse marker on the basis of the criterion
of compositionality, i.e. by virtue of the fact that it comprises part of
the phrase as you know. As you know, in its entirety, could be considered
a discourse marker, but in that case, the appropriate annotation would
be “as_Z4[i1.3.1 you_Z4[i1.3.2 know_Z4[i1.3.3” rather than “as_Z5
you_Z4[i1.2.1 know_Z4[i1.2.2”, where i1.3.1 marks the first segment in
an idiom comprising three lexical items, i1.2.2 tags the second segment
in an idiom that consists of two lexical items, etc.
On the other hand, there are lexical items that are invariably
tagged with the same (sometimes discourse marker-relevant and other
times non-discourse marker relevant) tags regardless of their syntactic
(non-)integration and functional scope.
An example for discourse marker-relevant invariant tagging is actually,
which might be used as a discourse marker that has the ensuing discourse
unit in its scope (14a) or as an adverbial modifier that has scope over the
verb it modifies as in 14b below (all extracts are from the USAS-tagged
CI corpus, emphases are mine):
(14a) No_Z4,_PUNC that_Z8 was_A3+ n’t_Z6 exactly_A4.2 +

the_Z5 reason_A2.2 ._PUNC Actually_A5.4 + ,_PUNC what_Z8
it_Z8 was_A3+ ,_PUNC is_Z5 I_Z8mf felt_X2.1 that_Z5 films_Q4.3
were_Z5 getting_A9+ they_Z8mfn started_T2+ to_Z5 be_Z5 repeat-
ing_N6+ ._PUNC
(14b) They_Z8mfn ‘re_A3+ one_T3 of_Z5 the_Z5 few_N5-
cats_L2 mfn in_Z5 the_Z5
world_W1 that_Z8 can_A7+ actually_A5.4 + swim_M4
under_M4[i619.2.1
water_M4[i619.2.2
24 P. B. Furkó
An example for non-discourse marker-relevant invariant tagging is now,

which can be used as a discourse marker that marks topic shift (15a)
or as a circumstance adverb (15b). However, USAS does not distinguish
between discourse marker and non-discourse marker uses of now, and
both are labelled as T1.1.2, i.e. as “general terms relating to a present
period/point in time”:
(15a) Good_Z4[i297.2.1 heavens_Z4[i297.2.2,_PUNC such_Z5

an_Z5 intelligent_X9.1+ man_S2.2 m is_Z5 excited_X5.2 +
about_Z5 a_Z5 movie_Q4.3 star_W1 ?_PUNC Now_T1.1.2
what_Z8 about_Z5 her_Z8f and_Z5 the_Z5 Kennedy_Z1mf ’s_Z5
?
(15b) Somebody_Z8mfc explain_Q2.2/A7 + to_Z5 Paris_Z2 and_Z5
Nicole_Z1f,_PUNC live_L1+ means_X4.2 we_Z8 ’re_A3+ on_Z5
television_Q4.3 right_T1.1.2[i7.2.1 now_T1.1.2 [i7.2.2 ._PUNC
Another discourse marker that displays “high frequency and versatility”

(Andersen 2001: 221) is like, whose functional spectrum ranges from
(16a) marking explanation/exemplification, (b) emphasis/new informa-
tion, to (c) quotative, (d) approximative and (e) softening functions. The
list of functions is far from exhaustive, for a more detailed functional
spectrum of like cf. e.g. Beeching (2016: 126ff ).
(16)
(a) My roommate never cleans when I ask him to. Like, I asked him
yesterday to clean, and he never did it. (Like_E2 + ,_PUNC I_Z8mf
asked_Q2.2 him_Z8m yesterday_T1.1.1 to_Z5 clean_B4,_PUNC
and_Z5 he_Z8m never_T1/Z6 did_A1.1.1 it_Z8 ._PUNC)
(b) This guy is so cool. I mean, he’s like the coolest person you could
meet. (I_Z4[i1.2.1 mean_Z4[i1.2.2,_PUNC he_Z8m s_T1.3 like_Z5
the_Z5 coolest_O4.6-person_S2mfc you_Z8mf could_A7+ meet_S3.1
._PUNC)
(c) I went to the clerk to ask him where the beer was, and he’s like,
‘I don’t know, I’m new here’, so I’m like, yeah, sure, like, you should
know this, man! (so_Z5 Im_Z99 like_Z5,_PUNC yeah_Z4,_PUNC
sure_A7+ ,_PUNC like_Z4,_PUNC
you_Z8mf should_S6+ know_X2.2+ this_Z8,_PUNC man_S2.2m)

(d) I missed like 40 questions on the exam. (I_Z8mf missed_A5.3-
like_Z5 40_N1 questions_Q2.2 on_Z5 the_Z5 exam_P1 ._PUNC)
(e) Could you, like, loan me $100? (Could_A7+ you_Z8mf,_PUNC
like_Z4,_PUNC loan_A9- me_Z8mf $100_Z99 ?_PUNC)
It is interesting to observe on the basis of the above examples that USAS

tags like as a discourse marker (i.e. Z4) in only one of its five main func-
tions: in (16a), it is incorrectly marked as “emotional state of liking”,
and in 16b, c and d, the tag Z5 indicates a fuzzy boundary between
the prepositional use (“grammatical bin”) and the discourse marker use
(“discourse bin”) of like: while the criteria of syntactic optionality and
non-propositionality would suggest that such uses are marked as Z4, the
grammatical interpretation (Z5) is also plausible in (b) and (d), where
no punctuation marks are used before and after like.
Yet another discourse marker that USAS has a hard time identifying
is the connective use of so. Syntactically integrated tokens receive either
A13.3 tags (“degree: boosters”), N5 tags (“quantities”) as in (17), or Z5
tags (“grammatical bin”), as in (18):
(17) but_Z5 they_Z8mfn have_Z5 wasted_N5.2+ so_A13.3

much_N5+ time_T1 and_Z5
so_N5 [i1.2.1 many_N5[i1.2.2 businesses_I2.1 have_Z5 gone_M1
bust_B1 as_Z5 a_Z5
result_A2.2
(18) But_Z5 why_A2.2 I_Z8mf feel_X2.1 so_Z5 frustrated_S8-
about_Z5 it_Z8 is_Z5
we_Z8 talked_Q2.1 about_Z5 this_Z8 at_Z5 the_Z5 start_T2 +
of_Z5 November_T1.3
._PUNC
The first token of so is most likely categorized as a subjective booster

(A13.3) on the basis of its co-occurrence with the word wasted, which
is associated with emotion (frustration), while so <A13.3> could also
be associated with quantification, similarly to the second token in the
26 P. B. Furkó
same utterance (tagged as N5). The indeterminacy between quantita-

tive and boosting functions of so reveals a fuzzy boundary between pri-
marily propositional (objective) and primarily evaluative/subjective uses
of the lexical items under scrutiny. Naturally, in the course of manual
annotation, the human annotator can have recourse to intonation and
other paralinguistic features of the utterance; however, it is still hard
to draw a boundary between clear-cut cases of non-boosting, quantita-
tive and, conversely, boosting, non-quantitative uses of so as the above
examples also illustrate. Moreover, it can be observed that in (18), so
also co-occurs with words associated with emotions (“feel” and “frustrat-
ed”, respectively), yet it is tagged as a purely grammatical item (Z5, i.e.
“grammatical bin”), rather than a booster.
Disambiguation between grammatical items and discourse markers
can also be an issue in utterances where so has a straightforward con-
nective function, connecting its host unit to the previous utterance and
marking a conclusive, resultative relation between the two discourse seg-
ments, as in (19):
(19) So_Z5 would_A7+ you_Z8mf bail_S8 +[i1.2.1 them_Z8mfn

out_S8+ [i1.2.2 in_Z5
the_Z5 short-term_T1.3- to_Z5 keep_A9+ those_Z5 jobs_I3.1
?_PUNC
Once again, USAS does not make a distinction between such uses and
grammatical uses exemplified by (18) above, marking so as Z5 (gram-
matical bin), rather than Z4 (discourse bin), the latter of which would
be more appropriate based on its syntactic detachment and connective
function.
Well, on the other hand, is a non-multi-word lexical unit whose dis-
course marker and non-discourse marker uses are clearly distinguishable
based on both manual and automated annotation, as (20) and (21) illus-
trate:
(20) But_Z5 that_Z8 goes_E2+[i1.2.1 for_E2+[i1.2.2 British_Z2

and_Z5 American_Z2
nuclear_Y1/G3[i2.2.1 weapons_Y1/G3[i2.2.2 as_N5++[i3.2.1

well_N5 ++[i3.2.2
._PUNC
(21) No-one_Z6/Z8cmf asked_Q2.2 them_Z8mfn,_PUNC “_PUNC
Well_A5.1 + what_Z8 does_Z5 that_Z8 mean_Q1.1 with_Z5
the_Z5 things_O2 you_Z8mf ‘re_Z5 not_Z6 going_T1.1.3[i1.2.1
to_T1.1.3[i1.2.2 go_T2++[i2.2.1 ahead_T2++[i2.2.2 with_Z5
?_PUNC “_PUNC
As the examples show, compositional/propositional uses are tagged

as N5 (quantities), especially in the phrase as well, while discourse
marker/interpersonal uses are marked as A5.1 (evaluation).
6 Conclusions, Utility of USAS

as a Heuristic Tool
In the above introduction, I argued that discourse markers are notori-
ously difficult to identify for humans and computers alike, and there are
several borderline phenomena, fuzzy boundaries and cases of ambiguity
resulting from discourse markers’ inherent, criterial features. In answer
to the research questions posed in Sect. 3 above, it can be observed that
the disambiguation methods automatic annotation uses are efficient for
filtering out non-discourse marker tokens of the most frequent discourse
marker types: thus, automatized annotation enables the researcher to
obtain an adequate global picture of the D-values of most of the lexi-
cal items that are frequently used as discourse marker types.
It was also demonstrated that the margin of error reported to apply in
general also applies to the identification of discourse markers collectively
and, in the case of multi-word units such as you know and I mean, indi-
vidually as well. However, we find a great degree of variation in the pre-
cision/margin of error with which non-multi-word discourse markers are
tagged. Such varying precisions are mostly due to discourse markers’ cri-
terial features of source category layering, syntactic non-integration, vari-
able/functional scope, all of which challenge the disambiguation meth-
ods USAS applies, with special reference to part-of-speech tagging, gen-
eral likelihood ranking and multi-word-expression extraction.
28 P. B. Furkó
7 Alternative Perspectives
on Discourse-Pragmatic Devices: Outline
of the Volume
While discourse markers will continue to puzzle humans and computers
alike, we can safely say that automatized methods can open new vistas
of research into the study of discourse markers, i.e. lexical items where
drawing a distinction between propositional and non-propositional, syn-
tactically semantically integrated and discourse-pragmatic uses is espe-
cially relevant. Moreover, discourse markers cannot be studied in isola-
tion, but with reference to other discourse-pragmatic devices (quotatives
in Chapters 2, 3, 8 and 9, contextualization cues in Chapter 3, modal
particles in Chapters 4 and 8 and pragmatic routines in Chapters 7 and
8), while the analysis of non-propositional items needs to be comple-
mented by the study of the contribution of propositional lexical items
that belong to related semantic fields (cf. Chapters 7–9). Secondly, only
a mixed methodology that involves computerized means as well as the
human element, a quantitative analysis of large datasets as well as careful
qualitative analysis of individual examples needs to be undertaken when
analysing genres/sub-genres (Chapter 2) or when reflecting critically on
discourses (Chapters 3 and 4), input for SLA (Chapter 5) as well as the
adequacy of translated texts (Chapters 6–9).
I hope to further illustrate these points in the ensuing chapters with
regard to genre analysis (Chapter 2) and other applied linguistic fields
such as Critical Discourse Analysis (Chapters 3 and 4), second language
acquisition (Chapter 5), translation theory (Chapters 6–9), the analysis
of scripted discourse in contemporary cinematography (Chapters 6 and
7), literary discourse (Chapter 8), as well as Bible studies (Chapter 9).
In the individual chapters, I use my previous research as a starting
point and reconsider previous findings from new methodological per-
spectives as well as new datasets, in the course of which the following
principles have been applied:
Another random document with
no related content on Scribd:
catalogue, which had been rehearsed so many times.
"There you go again, breaking right into the middle of a

sentence," said Grace. "What would your mother say?"
"Perhaps she would say, 'Don't be always lecturing the

child, Grace,'" said I mischievously, quoting some words I
had overheard from my mother.
Then, as I saw by her rising color that she was really

angry, I threw my arms round her and hugged her.
"There, don't be vexed, Gracy dear; you know I would

not disturb maman for the world. But I do really want to go
to the farm very much to teach Lucille the lace-stitch you
showed me yesterday, and to see the new kittens."
"Kittens! What kittens?" said Grace, who was a dear

lover of pussies of all sorts.
"Why, the new kittens. Don't you remember the

beautiful young cats that David brought to his mother the
last time he came home? One of them has kittens, and
Mother Jeanne says I may have my choice of them."
"Oh, yes; go by all means, my dear; and I hope you will

have a pleasant day. Only be sure you are at home before
dark, and mind you don't wait till it is time you were here
before you set out. And, as to the kitlings, if there should be
a tortoise-shell or a dark brindle, I would choose that,
especially if it have a white face. Such cats are always
good-tempered and good mousers."
"I believe these cats are all white," said I; "the mother
is as white as snow."
Grace's face was shadowed a little.

"I don't know about that," said she doubtfully. "In
Cornwall, we think that white cats bring ill-luck. My poor
sister had a beautiful white cat come to her, and that very
night she broke her china jug, and the next day her
husband fell from the tall pear-tree and was lamed for life."
"But these are not like common cats, you know," said I,
suppressing a laugh which I knew would mortally offend
Grace and perhaps lose me my holiday. "They are
outlandish cats, with long hair and bushy tails. I should
think that would be different."
"Perhaps so; but I would think about it a little. However,

I will come down and see them myself."
I tiptoed through my mother's room into my own little

cell, collected my working things into the pretty foreign
basket which David had brought me the last time he came
home, and then, kissing my mother's pale cheek, I
descended the stairs softly, and did not give a single skip till
I was beyond the precincts of the tower.
"How full of notions Grace is," I said to myself. "I

wonder if all the Cornish people are like that." * (N.B. † If a
hare had run across my own path, or I had heard a crow on
my left hand, I dare say I should have turned back from my
expedition.) "But I mean to have the kitten in spite of her.
As though I would give up a beautiful long-haired white cat
for such a fancy as that!"
* They are, even to this day.—L. S.
† N. B.—nota bene
I did not hasten on my way, for it was early, and I found
my walk so pleasant that I had no desire to shorten it. The
bramble-berries and filberts that were ripening by the sides
of the lane had great attractions for me. There were late
autumn flowers to gather, and lizards to watch as they ran
to and fro on the walls or sunned their gilded sides on a
broad flat stone, vanishing like a shadow when one drew
near. A great wind had blown the day before and thrown
down many apples from the trees that overhung the lane.
I filled my pocket with some ripe golden pippins, and

walked on eating one till I drew near the place where the
highway to Avranches, such as it was, crossed our lane.
This was a favorite resting-place, since it commanded a
glorious view of sea and shore and the great fortress-
monastery. There was a kind of crag or projecting rock
some thirty feet high, round which the road wound, and
which, while it presented a perpendicular face to the
highway, was easily ascended by an active person from the
side of the lane.
"I wonder whether they are gathering the vraic," I said

to myself. "I should think a great quantity must have come
ashore after the wind last night. I mean to climb up and
see." *
* The vraic or varech is the seaweed, which is very

abundant on this coast, and much esteemed for manure.
It is regularly harvested in spring and autumn, but may
be gathered at any time.
I climbed lightly up the rude rocky steps, but started as

I came upon Lucille, who was sitting upon the dry moss
which covered like a soft carpet the top of the rock. She
was wrapped closely in her long black cloak, the hood of
which was drawn over her head, somewhat to the detriment
of her clean starched cap. Her unfailing companion, the
distaff, was in her girdle, but the spindle lay idle beside her,
though she seemed to have cleared a flat place especially
for it to dance upon. Her hands were folded over her knee,
and her eyes were fixed upon the high road, which from this
elevated point could be traced all the way to Avranches.
I saw in a moment that she was in one of her moods,

but I was in too high spirits with my walk and my holiday to
mind that. And as she did not seem to hear my approach, I
put my two hands over her eyes, saying, in the words of our
child's game, "Guess whose fingers are all these."
"Vevette, how you startled me!" she exclaimed, rather

angrily. And then, recovering herself, "How did you come
here?"
"On my feet, since I have no wings," I answered, sitting

down beside her on the dry moss. "Maman gave me a whole
holiday because she has a headache, and I thought I would
come down and teach you my new lace stitches. It is well I
took a fancy to climb up here, or I should have missed you.
But now, tell me how came you here?"
"Because I have a holiday as well as yourself,"

answered Lucille, in a tone which had no pleasure in it.
"Aunt Denise has come up from Granville to see my mother,
and maman said I might have a play-day too, and go to see
Marie Lebrun if I liked. But I don't care about going. I know
they only sent me away because they have secrets to talk
about which they don't want me to hear."
"Well, why need you mind?" I asked. "Maman often says

to me, 'Run away, petite, I wish to say something to Grace,'
and I never mind it a bit. Of course grown people have
things to talk about which they don't want children to hear.
Why should you care?"
"But I do care," said Lucille, and her eyes with tears. "I
am not a child like you. I am three years older, and I do
think they might trust me."
"It is not that they do not trust you, silly one," I

returned, a little out of patience with the mood I could not
comprehend. "As I tell you, there are things to be talked
about by grown people which girls do not understand and
ought not to know. Mrs. Grace has told me that a dozen
times. What is the use of minding? We don't understand,
and there is the end. Some time we shall, I suppose."
Lucille did not answer. She fixed her eyes once more on
the highway, and I let mine wander off over the sands and
the shore where people, looking like little black ants, were
busily collecting the precious seaweed, to Mount St.
Michael, whose turrets shone brightly in the sun.
"I wish I had wings," said I at last. "How I should like to

fly over the sands and alight on the top of the mount
yonder, where the great gilded angel used to stand looking
over land and seas. I wonder whether he got tired of his
perch and flew away some night."
"You should not speak so of the holy angels. It is not

right," said Lucille gravely.
"I was not speaking of the angel, but of his image," said
I; "that is quite another thing. Then I would spread my
wings and travel over to the islands yonder, and then to
England, where my uncles live."
"And get shot for a strange water-fowl," said Lucille,

apparently diverted for the moment, and laughing at my
fancy. "Then you would be stuffed and set up to be gazed at
for sixpence a head, and that would be more tiresome than
sitting at your embroidery."
"Yes, I don't think I should like it at all. Let me take the

distaff, Lucille. I have not spun any thread in a long time.
What beautiful fine flax!"
"Yes, it is some that my aunt brought me. She got it of

a ship-captain who came from foreign parts. Take care you
don't break my thread."
We chatted on indifferent subjects a while, and Lucille

seemed to have recovered her good humor, when I
inadvertently disturbed it again.
"Martin said he met your father coming from Avranches

yesterday. What took him so far from home?"
"I don't know; they never tell me anything," answered

Lucille, her face clouding.
"There might be a very good reason for his not telling

you," I remarked in a low tone. "If his journey was about
the Religion, it might be a great deal better for you to be
able to say you did not know. And I dare say it was, for my
father has been away a great deal of late."
"Oh, the Religion—always the Religion!" said Lucille

between her teeth; "I hate the very name of the Religion."
"Lucille, how dare you?" I gasped, rather than spoke. I

was too shocked to say more.
"Well, I do," she returned vehemently. "It spoils

everything. It separates families and neighbors, shuts us up
just to our own little selves, and cuts us off from everything
that is pleasant. Jennette Maury can go to the Sunday fêtes
and the dances on feast days under the great chestnut, but
I must stay at home and read a musty book, because I am
of the Religion. Other people live in peace, and nobody
interferes with them. We live with a sword hung over our
heads, and our daily path is like that over the Grève yonder
—likely to swallow us up any time. And what do we gain by
it in this world, I should like to know?"
"What should we lose in the next world if we deserted

it?" I asked, finding my voice at last.
"I am not talking of deserting it. I am no Judas, though

they seem to think I am by the way they treat me—never
telling me anything. But I don't see why we should not have
kept to the ways of our fathers, and saved all this trouble."
"WE DO keep to the faith of our fathers," said I,

repeating the proud boast of the Vaudois, which I had long
ago learned by heart. "Our church never was corrupted by
Rome, and did not need reforming. But, Lucille, what would
your father and mother say to such words?"
"I should never say such words to them," answered

Lucille, "and I am foolish to say them to you. I suppose,
however, you will go and repeat them to every one, and let
the world say how much better and more religious is the
heiress of the Tour d'Antin than poor Lucille Sablot."
"Lucille, you know better," I answered indignantly; "but

I see you don't want anything of me, so I shall go home
again, as you say Mother Jeanne is busy."
And gathering up my basket and laying down the distaff

in Lucille's lap, I rose to depart, though I trembled so much
with excitement and indignation that I could hardly stand.
Lucille looked at me in surprise, for in our ordinary
quarrels, I grew cool as she grew angry, and vice versâ.
"Don't go, Vevette. I ought not to have spoken so. I did

not half mean it, but I am so very, very unhappy."
As she spoke, she hid her face and burst into a flood of
tears and sobs.
I sat down again, knowing from experience that when

she recovered from her crying fit, her bad mood would be
gone for that day.
So it proved. After sobbing a long time, she wiped her

eyes and made a great effort to compose herself.
"I am sorry I was so cross," said she; "but I am so

unhappy. There is so much that I cannot understand. Why
should you be the heiress of d'Antin and I only a poor
farmer's daughter? Why should you learn music and English
and dress in silk, while I wear homespun and tend sheep,
and come and go at everybody's call? Why should our
enemies triumph and eat us up like bread, and live in all
sorts of luxury, while we are poor and trodden down like the
mire in the streets, and our Master never put forth a hand
to help us? We give up everything for him, and he lets us be
beaten on every side, and gives us nothing but promises—
promises for another world, from which nobody has come
back to tell us anything. No, I don't understand it."
Lucille spoke with a fire and passion compared to which

her former vehemence was nothing.
I had never thought of these things—never dreamed of

questioning anything that was taught me. Indeed, I believe
I had been too full of dreams to think at all. I was stricken
dumb before her at first, but as she gazed at me with her
dark eyes like sombre flames, I felt I must say something,
so I gave the only answer that occurred to me—the only
one indeed that I have ever found.
"It is the will of God, Lucille, and he must know best."
Lucille muttered something which I did not quite hear.
"And besides, he does help us," I added, gathering

courage. "Just think how all the martyrs have been helped
to stand firm, and what joys they have felt even at the
galleys and in dark dungeons, where they had hardly room
to breathe."
"I know they say so," said Lucille; "but tell me, Vevette,
have you experienced any of these wonderful joys. Because
I know I never did."
I did not know exactly what to answer to this question.

In fact, in those days my conscience was in that uneasy
state in which it always must be with any half-hearted
person. No, I could not say that my religion was any
comfort to me, and I hastened to change the conversation.
"Anyhow, Lucille, I don't think you would be any happier

if we were to change places. You would be lectured and
ordered about, and sent out of the way a great deal more
than you are now, and you would not have nearly as much
time to yourself. I believe, after all, it is more in being
contented than anything else. Look at Gran'mère Luchon.
She has as little as any one I know—living down by the
shore in that dark smoky little hut with her two little
grandchildren, and supporting them and herself with her
net-making and mending and her spinning. And yet she is
happy. She is always singing over her work, and I never
heard her make a complaint."
"She is not there any more," said Lucille. "The new curé
ordered her to go to mass, and because she would not, he
has taken the children away and handed them over to the
nuns, and nobody knows what, has become of the old
woman."
"The wretches!" I exclaimed.
"Hush!" said Lucille. "Don't speak so loud; nobody

knows who may be listening. I hate living so—in such
constraint and danger all the time. It is odious."
"Don't let us talk about it any more," said I. "I have

some news for you. My cousin, Andrew Corbet, from
England, is coming to visit us. Will it not seem odd to have
a cousin?"
"Not to me," said Lucille, making an effort to throw off

her moodiness. "I have a plenty of them, you know. When
do you expect him?"
"Next week, perhaps; the time is not set."
"What is he like?"
"I don't know; I have never seen him. He is about

twenty years old, and has been educated at a great college
in England, so I suppose he is like other young gentlemen.
Come, let us eat some of Mrs. Grace's cakes and bonbons,
and then I will show you my new stitch. Grace gave me a
nice basket, because she said we might like to make a little
feast under the trees."
Lucille had something too—a bottle of milk and some

wheaten bread which she had set out to carry to Gran'mère
Luchon, when she heard of the misfortune which had
befallen the poor woman. We grew quite merry over our
little feast, and the lesson in needlework went on
prosperously afterward.
"You have caught it beautifully," said I. "Mrs. Grace

would say that you excelled your pattern. But what are you
looking at?"
For Lucille had dropped her work and was gazing

intently in the direction of Avranches.
I turned my eyes the same way and beheld a procession

coming up the road—of what sort I could not at first
discover. There was a cross-bearer and two or three
banners; then a sight dreaded by every Huguenot child in
France—the Host carried under a fine canopy—and then
came a dozen or so of donkeys, each led by a man and
bearing a woman dressed in black, with a white scapular
and long black veil.
"They are the nuns coming to take possession of the

hospital," said Lucille. "It has been all repaired and fitted up
anew, and they are to have a school and teach lace-making
and embroidery."
"Lucille, what do you mean?" I exclaimed; for she had

risen and stepped to the edge of the rock to have a better
view. "They will see you. Come down here behind the
bushes till they are past."
Lucille obeyed rather unwillingly, as I thought.
We peeped through the bushes as the procession

advanced, and had a good view of the nuns. There were ten
of them, riding with eyes cast down and hands folded in
their large sleeves. One or two of them were very pretty,
and all had a ladylike look.
Last came the two little grandchildren of poor Mère
Luchon. The youngest, a mere baby, was sucking a lump of
gingerbread, apparently quite content; but the sobs and
tear-stained face of the other told a different story. She was
seven years old, and was already a great help and comfort
to the old woman. As she passed, she raised her streaming
eyes as if imploring pity.
My blood boiled at the sight, and if I could have

commanded the lightning from heaven, that procession
would have gone no farther. It was closed by a number of
villagers, all telling their beads, some with a great show of
devotion, others languidly and carelessly enough.
The new curé came last of all. He was a small, thin,

sharp-faced man, with a cruel mouth, and eyes that seemed
to see everything at once. He was certainly a great contrast
to poor Father Jean, who used to go about with his deep
pockets filled with bonbons, which he distributed to Catholic
and Protestant children alike.
"The wretches! The murderous brigands!" said I

between my teeth. "Oh, if I could kill them all! The vile
kidnappers! Oh, why does the Lord suffer such things?"
"That is what I ask," said Lucille. "Why should they be

so prospered and have so much power if the Lord is not on
their side? As to these children, I don't know that I pity
them so very much. The old woman could not have lived
long, and now they are sure of support and a good
education. I think the nuns are very kind-looking ladies, for
my part. And if they were right after all—if one's salvation
does depend upon being a Roman Catholic—then they are
right in forcing people to become so."
"Why did not our Lord and his apostles force all the
Jews to become Christians?" I demanded hotly enough. "He
said he had only to ask to receive more than twelve legions
of angels. Why did not he do it, and shut up all those people
who did not believe on him, or put them to death, if that is
the right way?"
"He said his kingdom was not of this world, else would
his servants fight," answered Lucille.
"Then the kingdom which is of this world, and whose

servants do fight and oppress, is not his," I answered, for I
could reason well enough when I was roused from my
daydreams.
"We ought to be going," said Lucille, abruptly changing

the subject. "The supper will be ready, and my father will be
angry if I am not there. I am to be kept to rules as if I were
no more than five years old."
Jeanne welcomed me with her usual affection, but her

eyes were red with weeping, and she was evidently absent-
minded.
I told her what we had seen.
"Yes, I have had the story from my sister," said Jeanne,

her eyes overflowing as she spoke. "The poor old woman!
Happily it cannot be long in the course of nature before she
goes to her rest, but my heart aches for the little ones. My
children, you must be doubly careful. This new priest is not
like the old one—he will leave us no peace. You must take
care never even to go near the church, or stop to look on at
any of their doings. Perhaps a way of escape may be
opened to us before long. It would indeed be hard to leave
our home and go among strangers, but exile with liberty of
worship would be better than living in such constant fear."
"Put thy trust in God, my Jeanne," said Father Simon.
"We are all in his hands. We must remember that the
church has never been promised anything in this world but
tribulation and the cross. The crown is to come hereafter.
Now let us think of something else. Mamselle Vevette, will
you come and help to gather the apples on your own tree?
They are quite ready, and I will carry them up for you when
you go home."
I had been grave quite as long as I liked, and was very

ready to enjoy the apple-picking from my own particular
tree of golden Jeannetons, which had been solemnly
planted when I was born, and now hung loaded with fruit.
Never were such apples as those, I am sure. I wonder
whether the tree is still in bearing? It must be old and
moss-grown by this time, if it has not been cut down.
Jeanne made us a supper of fresh pan-cakes, galette,

fruit, and rich cream cheese, and when I went home, Father
Simon shouldered his hotte * and carried a famous load of
beautiful apples up to the tower.
* A kind of deep, roomy basket, made to be carried on

the shoulders.
I found my mother much better, and able to welcome

me, and to hear all I had to tell her. I hesitated about
repeating my conversation with Lucille on the rock, but my
mind had been so disturbed that at last I thought best to do
so, hoping to have my doubts laid at rest.
"You gave the right answer, my little one," said my

mother when I had finished. "It is the will of God.
Remember that he has never promised his children
temporal prosperity. 'In the world ye shall have tribulation,'
are his own words. Yet he does give his children many
pleasures. There are beautiful flowers and fair fruits
growing even by the side of the strait and narrow way, but
we must not go out of the way to seek them. Neither must
we be discouraged when the path leads over rocks and
thorns, or even through marshes and quicksands; but
remember that our dear Lord has trodden every step before
us, and is waiting to receive us at the end."
Much more she said, in the same wise and gentle strain,
and at last sent me to bed feeling somewhat comforted. The
night was warm, and my door was left ajar for air. I had
hardly fallen asleep, as it seemed to me, when I was waked
by voices, and heard my mother say:
"I do not like what she says about Lucille. I fear the girl
has been tampered with. Perhaps we should warn her
parents."
"We will think about that," said my father. "Ah, my

Marguerite, if you and the little one were but in safety—"
"Do not ask me to leave you, Armand—not yet," said

my mother, clasping her hands. "If we could but send the
child home to my sister, I should be at ease. Could we not
do it, when Andrew comes?"
"We will consider of it," answered my father. "And now,

my Pearl, let us betake ourselves to prayer."
The murmured sound of the prayer sent me to sleep,

and I heard no more, but I turned Lucille's words over in
my mind with a vague uneasiness many times during the
next few days. I was destined to remember them for long
afterward.
The next day was made memorable by an unlucky
accident. Mrs. Grace was standing in the door of my room
(which I have said was raised several steps), lecturing me
in her usual prim fashion concerning certain untidinesses
which she had discovered about my toilette-table, when,
suddenly stepping backward, she fell down the stairs,
bruising herself and spraining her ankle very badly.
We dared not send for a surgeon. There was an old man

at Avranches who was very skilful, and with whom we had
always been on good terms, though he was a Roman
Catholic; but he had lately taken a young assistant (or
rather had been given one, for we all believed the young
man had been placed as a spy over the old one), and
should it be known that we had a sick person in the house,
we were in danger of being invaded by the priests, striving
to force or coax the sick person into a recantation.
Happily my father had a pretty good practical

knowledge of surgery, and both my mother and Mrs. Grace
herself were strong in the virtues and uses of herbs and
simples.
Mrs. Grace was presently put to bed and her ankle

bandaged. She was in great pain, but the pain was little or
nothing compared to the worry of helplessness,
housekeeping cares, and the necessity of being waited upon
instead of waiting upon others. Truth to say, she was but a
troublesome charge.
My dear mother, who had borne this same cross of

helplessness for many a year, preached patience in her
gentle way.
Mrs. Grace assented to all she said, called herself a

miserable, rebellious sinner, and the next minute fretted
more than ever: over that careless Marie, who would be
sure to burn the marmalade, or that stupid coward of a
Julienne, who would not venture up to the top of the tower
to bring in the drying fruit lest she should see the white
chevalier. For after a long season of absence—for what
ghostly purpose, who should say?—the white chevalier had
again been seen walking on the battlements of the round
tower, or passing the window of his wretched and guilty
wife's apartment.
"Do not trouble yourself about the marmalade, my poor

Grace," said my mother, with a somewhat woeful smile.
"Who knows whether we shall be alive to eat it, or whether
all our stores may not fall into the hands of our enemies?"
"I should like to spice the marmalade for them!"

exclaimed Grace, quite overcome by the idea of her dainties
being devoured by the Papisties, as she always called them.
"And as to the tower," continued my mother, "I think

myself the maids may as well keep away from it. If the
white chevalier and his wife should really have been seen, it
is just as well not to run any risks."
"But whom then will you trust?" asked Grace, with a

startled look.
My mother put to her lips a fresh rose she had brought

in her hand, and glanced at me, and Grace said no more. I
was not annoyed, as Lucille would have been, for I had
become accustomed to such hints; and with a passing
wonder as to whether my mother really believed in the
white chevalier, I plunged into my dear "Arcadia," and
forgot all earthly cares in the somewhat long-winded trials
of the virtuous Parthenia. But I was destined to hear more
of the matter.
That very evening, about an hour before sunset, my
father asked me to walk with him. This was a great honor,
for in my youth, children were by no means so familiar with
their parents as they are now. Whether the change be for
the better or no depends upon the parents a good deal.
We walked out by the lane, across a field, and through

the loaded orchard bending with golden and ruddy fruit,
some of which was already gathered for the cider-mill. The
low sun shone under the branches, and turned the heaps of
apples to heaps of gold and rubies. It was very still, but the
tide was high, and came in over the distant sands with a
hollow roar, which my father said portended a storm. He
spoke little till we reached a little heathy eminence crowned
with one of the monuments of ancient date so common in
Normandy and Brittany. From this point we had a view for a
long distance around, and nobody could come near us
without being observed. My father sat down on one of the
fallen stones, and motioned me to sit beside him.
"My daughter," said he, taking my hand in his with a

certain solemnity, "you are now almost a woman, and old
enough to be admitted into the knowledge of your father's
secrets. But such knowledge is full of danger. Are you
brave, my child? Are you a worthy descendant of those
valiant Provençal and Vaudois women who hazarded their
lives for the faith? Consider, my Vevette! Suppose you were
required to go into the upper floor of the old tower, even to
the ladies' bower, at night; would you be afraid to do it?
Consider, and give me an answer."
All my better self rose up at this appeal. I considered a

moment, and then answered firmly—
"I might be afraid, but I would do it, if it were my duty."

"There spoke a true Corbet woman!" said my father,
smiling kindly on me and pressing the hand which he held.
"'MY DUTY!' Let that be your motto, as it is that of your
mother's house, and you will not go far wrong. Now listen
while I impart to you a weighty secret. But let us first make
sure that there are no eavesdroppers."
My father raised himself from the fallen stone and

looked all around, but no one was in sight, and the sparse
heath and short grass could not hide anything so large as a
child of a year old. He even parted the brambles and wild
vines and looked inside the monument (which was one of
those made of three upright stones with a slab laid over the
top), but found nothing worse than a pair of young owls and
their mother, which were terribly disconcerted by his
scrutiny, and hissed and snapped valiantly.
Meantime I waited with anxious curiosity, though I had

a guess of what was coming.
"I have certain intelligence," said he, speaking in a low

voice, "that one of our best and oldest pastors, Monsieur
Bertheau, who has, at the risk of his life, visited and
comforted many of our afflicted brethren in Charenton and
elsewhere, is now flying from his enemies, and will arrive at
this place some time to-night. He must be lodged in the old
tower till the period of spring tides, when I shall hope to
procure a passage for him to Jersey, or to England itself.
Grace, who has usually taken charge of such fugitives, is
now disabled. I must be away this night, and your mother is
unable to do what is needful; besides that, her absence
from her room might excite suspicion. Mathew grows old
and forgetful, and I dare not trust any of the other
servants. Dare you, my daughter, undertake to meet this
venerable man in the ruins of the chapel to-night, and lead
him by the secret passage to the room at the top of the
tower, which has been prepared for him?"
"Yes, my father," I answered; "but how shall I know the

way?"
"I will give you directions which will lead you to the
entrance of the passage. Turn to your right after that, and
you cannot miss your way. When the good man is in safety,
you can come directly to your mother's room by another
passage, which I will also indicate to you. But, my child, I
must not conceal from you that there is danger in this trust.
Should you be discovered by any of our enemies in giving
help to this good old man, your life or your liberty must be
the forfeit."
"I know it, my father," I answered; "but if it is my duty,

I can do it. Besides, there is danger anyhow."
"That is true, my child. He that saveth his life is as like

to lose it as he that layeth it down for the Lord's sake and
the Gospels."
Then my father broke down, clasped me in his arms,

and wept over me in the way that is so terrible to see in a
strong man.
"My child, my Marguerite's only child! My treasure! And

must I lay down thy young life also? Oh, Lord, how long,
how long!"
Presently, however, he composed himself, and laying his

hand on my head, he most solemnly dedicated me to God
and his service, as the most precious thing he had to give.
That dedication has never ceased to affect my life, even
when I have strayed the farthest.

Discourse Markers and Beyond Descriptive and Critical Perspectives On Discourse Pragmatic Devices Across Genres and Languages 1St Ed 2020 Edition Peter B Furko Full Chapter PDF

Uploaded by

Copyright:

Available Formats

You might also like

Discourse Markers and Beyond Descriptive and Critical Perspectives On Discourse Pragmatic Devices Across Genres and Languages 1St Ed 2020 Edition Peter B Furko Full Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Discourse Markers and Beyond Descriptive and Critical Perspectives On Discourse Pragmatic Devices Across Genres and Languages 1St Ed 2020 Edition Peter B Furko Full Chapter PDF

Uploaded by

Copyright:

Available Formats

Discourse Markers and Beyond:

Descriptive and Critical Perspectives

A Contrastive View of Discourse Markers: Discourse

Discourse Markers: An Enunciative Approach 1st Edition

Language, Vernacular Discourse and Nationalisms 1st ed.

Ethical Discourse in Finance: Interdisciplinary and

Foreign Languages in Advertising: Linguistic and

Expressions of War in Australia and the Pacific:

The Politics of Climate Change Metaphors in the U.S.

Environmental Organizations and Reasoned Discourse

More information about this series at

Postdisciplinary Studies in Discourse

© The Editor(s) (if applicable) and The Author(s) 2020

Cover illustration: RooM the Agency/Alamy Stock Photo

1 Preliminary Issues: Category Membership, Methodology,

Part I Discourse Markers Across Genres

2 Discourse Markers in Natural Conversations,

3 Discourse Markers from a Critical Perspective:

3 The Political News Interview as a Genre 71

4 Discourse Markers from a Critical Perspective:

Part II Discourse Markers Across Languages

5 The Use of Discourse Markers in Business English

6 Discourse Markers in Scripted Discourse I: Issues

3.2 The Functional Spectrum of I Mean 151

7 Discourse Markers in Scripted Discourse II:

8 Discourse Markers and Their Translation in Literary

9 The Use of Discourse Markers in Bible Translations 219

2 Greek Sentence Conjunctions in John and Galatians

Concluding Remarks 247

Software and Online Tools 251

Data Availability 253

Appendix—List of Abbreviations 255

1 Theoretical and Empirical Approaches

issues include the lack of generally accepted terminology and classifica-

2 Characteristics of Discourse Markers,

Many scholars (cf. Schourup 1999) consider non-propositionality (non-

While in addition is indeed truth-conditional, the above test would pre-

Discourse markers’ extreme context-dependence is frequently identified

In addition to playing a role in pragmatic inferencing, individual dis-

information, and so on (cf. e.g. Beeching 2016: 4ff ). Discourse mark-

2.4 Weak Clause Association and Variable Scope

It is frequently observed in the literature that discourse markers usually

Conjuncts are more like disjuncts than adjuncts in having a relatively

In addition, some of the items that Quirk et al. refer to as “disjuncts”

(4a) Ed and Doris loved each other.

(5) My husband got a notice t’go into the service

Schiffrin (1987) concludes that and has “freedom of scope”, rather

2.5 Procedural Meaning/Non-compositionality

Although most scholars treat non-compositionality as a property of dis-

(6a) Seriously, you will have to leave.

A second test Blakemore uses is to see if a given item can combine

2.6 High Frequency, Orality, Stigmatization

However, it is easy to illustrate the meaningfulness and distinctive (as

In (9a) Original, you know comments on what is meant by “look right”,

3 Automatic Semantic Annotation:

evaluations of the accuracy of the system reported a precision value of

4 Corpus and Methodology

• a corpus of the official transcripts of 39 confrontational type of medi-