Variation and Change

Handbook of Pragmatics Highlights (HoPH)

The ten volumes of Handbook of Pragmatics Highlights focus on the
most salient topics in the field of pragmatics, thus dividing its wide
interdisciplinary spectrum
in a transparent and manageable way. Each volume starts with an upto-date overview of its field of interest and brings together some 1220
entries on its most pertinent aspects.
Since 1995 the Handbook of Pragmatics (HoP) and the HoP Online (in
conjunction with the Bibliography of Pragmatics Online) have provided
continuously updated state-of-the-art information for students and
researchers interested in the science of language in use. Their value as a
basic reference tool is now enhanced with the publication of a topically
organized series of paperbacks presenting HoP Highlights.
Whether your interests are predominantly philosophical, cognitive,
grammatical, social, cultural, variational, interactional, or discursive, the
HoP Highlights volumes make sure you always have the most relevant
encyclopedic articles at your fingertips.

Editors
Jef Verschueren
University of Antwerp

Jan-Ola stman
University of Helsinki

Volume 6
Variation and Change. Pragmatic perspectives
Edited by Mirjam Fried, Jan-Ola stman and Jef Verschueren

Variation and Change

Pragmatic perspectives
Edited by

Mirjam Fried
Institute for the Czech Language,
Czech Academy of Sciences, Prague

Jan-Ola stman
University of Helsinki

Jef Verschueren
University of Antwerp

John Benjamins Publishing Company

Amsterdam/Philadelphia

TM

The paper used in this publication meets the minimum requirements

of American National Standard for Information Sciences Permanence
of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data

Variation and change : pragmatic perspectives / edited by Mirjam Fried, Jan-Ola stman and Jef
Verschueren.
p. cm. (Handbook of Pragmatics Highlights, issn 1877-654X ; v. 6)
Includes bibliographical references and index.
1. Language and languages--Variation. 2. Linguistic change. 3. Pragmatics. I. Fried, Mirjam. II.
stman, Jan-Ola. III. Verschueren, Jef. IV. Title. V. Series.
P99.4.P72V364 2010
417.7--dc22 
2010036643
isbn 978 90 272 0783 8 (Pb ; alk. paper)
isbn 978 90 272 8917 9 (Eb)

2010 John Benjamins B.V.

No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other
means, without written permission from the publisher.
John Benjamins Publishing Co. P.O. Box 36224 1020 me Amsterdam The Netherlands
John Benjamins North America P.O. Box 27519 Philadelphia pa 19118-0519 usa

Table of contents

Introduction: From instances of change to explanations of change

Mirjam Fried
1.
Approaches to variation and change 2
2. Diachronic analysis and pathways of change 5
3.
Intra-lingual variation 8
4. Cross-language variation 10
5.
Summary and outlook 14

Borrowing
Jeanine Treffers-Daller
1.
Introduction 17
2. Short historical overview 18
3.
Definitions of borrowing: Terminological issues 19
4. Different types/classifications of borrowing 21
5.
The integration of borrowings 23
6. Constraints 25
7. Quantitative approaches 29
8. Psycholinguistic approaches 30

17

Contact linguistics
Michael Meeuwis & Jan-Ola stman
1.
Introduction 36
2. Contact in relation to classificatory bases in linguistics 36
3.
Contact and location 37
4. Direction of interference 38
5.
Indirect influence in language contact 42
6. Contact as process: Towards pragmatics 42

36

Creoles and creolization

Salikoko Mufwene

46

Dialect
Ronald Macaulay
1.
Introduction 61

61

VI

Variation and Change

2.

3.
4.
5.

 ther labels 62
O
2.1 Variety 62
2.2 Lect 63
2.3 Vernacular 63
2.4 Social dialects 64
Dialects as local forms of speech 64
Some examples of dialect studies 67
Some recent developments in the U.S. 68

Dialectology
Georges De Schutter
1.
Definition 73
2. An outline of history 73
3.
Aims of the investigation 75
3.1 The impact of history on the origin and evolution
of languages 76
3.2 Linguistic reconstruction 76
3.3 The study of universals of language,
especially implicational scales 77
3.4 Sociopragmatic and attitudinal aspects 77
3.5 Communicative aspects 78
4. Dialect atlases 79

73

Evolutionary pragmatics
Wolfgang Wildgen
1.
Pragmaticism, pragmatics, adaptation and the evolution
of language 81
2. Sign-functions and their evolutionary significance 83
2.1 The triad of sign functions 83
2.2 The evolutionary interpretation of the triad of functions 84
2.3 Selective value of communication and symbolic behavior 86
3.
Can the pragmatics of tool production and tool-use tell us something
about the origin of language? 87
3.1 Instrumentality in higher mammals and man 88
3.2 Is tool-making a pragmatic source of propositional semantics? 91
3.3 Cro-Magnon life space and the pragmatic space
of decorated caves 93
4. From ecological to cultural pragmatics 95
5.
Conclusions 96

81

Table of contents

Historical linguistics
Louis Goossens
1.
Introduction 100
2. HL in pre-generative work 100
3.
HL and generative grammar 103
4. New perspectives 104
5.
Explicit 107

100

Historical pragmatics
Andreas H. Jucker
1.
Introduction 110
2. Origins 111
3.
Data problems 111
4. Topics 113
4.1 Discourse markers 113
4.2 Speech acts 114
4.3 Politeness 116
5.
New perspectives 118

110

Implicature and language change

Kate Kearns
1.
Introduction 123
2. Analyses citing conversational principles 124
3.
Implicature and metaphor 131
4. Implicature, metonymy and merger 134
5.
Inferences based on lexical concepts 135
6. Summary 139

123

Interlanguage pragmatics
Gabriele Kasper
1.
Definition and scope 141
2. Pragmatic comprehension 141
2.1 Comprehension of nonliteral utterances 141
2.2 Assessment of politeness 142
2.3 Sociopragmatic assessment 142
3.
Production of linguistic action 143
4. Development of pragmatic competence 144
4.1 Cross-sectional studies 144
4.2 Longitudinal studies 145
4.3 Theoretical accounts 145

141

VII

VIII Variation and Change

 ragmatic transfer 146

P
5.1 Definition 146
5.2 Positive transfer 146
5.3 Negative transfer 147
5.4 Transferability 148
6. Communicative effect 149
7. Pragmatic norms 150
8. Language teaching 151
9. Research methods 152
10. Further reading 153
5.

Jargon
Luisa Martn Rojo
1.
Introduction 155
2. Studies of delinquent jargon and the introduction
of a value-giving measure 156
3.
Interpretations of antilanguages and their functions 160
3.1 Halliday: The concept of antilanguage 160
3.2 Sociolinguistic functions of delinquent jargon 162
4. Varieties of jargon 165
5.
Conclusions 168

155

Language change
Raymond Hickey
1.
Introduction 171
2. Issues in language change 173
2.1 Internal and external factors 173
2.2 Simplicity and symmetry 174
2.3 Iconicity and indexicality 174
2.4 Markedness and naturalness 175
2.5 Telic changes and epiphenomena 176
2.6 Mergers and distinctions 177
2.7 Possible changes 178
2.8 Unidirectionality of change 178
2.9 Ebb and flow 179
3.
Change and levels of language 180
3.1 Phonological change 180
3.2 Morphological change 181
3.3 Syntactic change 182
4. The study of universal grammar 183
4.1 The principles and parameters model 184

171

Table of contents

S emantic change 186

Pragmatic change 187
Methodologies 188
7.1 Comparative method 189
7.2 Internal reconstruction 190
7.3 Analogy 190
8. Sociolinguistic investigations 190
8.1 Data collection methods 192
8.2 Genre variation and stylistics 192
9. Pathways of change 192
9.1 Long-term change: Grammaticalization 192
9.2 Large-scale changes: The typological perspective 193
10. Contact accounts 194
11. Language areas (Sprachbnde) 197
12. Conclusion 198
5.
6.
7.

Language contact
Yaron Matras
1.
Introduction 203
2. Societal multilingualism 203
3.
The bilingual individual 204
4. Language mixing in conversation 205
5.
Contact-induced language change 207
6. Contact, typology and language classification 210
7. Concluding remarks 212

203

Reconstruction
Derek Nurse

215

Register
Norbert Dittmar
1.
History of the term register 221
2. Systematization of the term 224
3.
Perspectives 228
3.1 Functional linguistics la Ferguson 229
3.2 Intercultural communication 230
3.3 Linguistic pragmatics 230
3.4 Variation linguistics 231

221

Typology
Bernard Comrie

234

IX

Variation and Change

Variational pragmatics
Klaus P. Schneider
1.
Introduction 239
2. Pragmatic coordinates 240
2.1 The scope of pragmatics 240
2.2 The place in pragmatics 241
2.3 Levels of pragmatic analysis 244
3.
Variational coordinates 246
3.1 Variation and identity 246
3.2 Pragmatic variation 249
3.3 What is a pragmatic variable? 250
4. Methodological concerns 252
4.1 Some basic principles 252
4.2 Data 253
5.
Development 254
5.1 History 254
5.2 Research trends 256
6. Perspectives 260

239

Index

268

Introduction
From instances of change to explanations of change
Mirjam Fried

Czech Academy of Sciences, Prague

The notion of variation can be defined in a number of ways and the content of this
volume reflects quite directly the multi-layered scope of topics, goals, and methods
that constitute the full range of research devoted to language variation and change.
Major areas represented in the contributions and commented on in this introduction
could be organized along the following general themes: analytic and methodological
approaches, which can be framed in various theoretical backgrounds or determined by
a particular set of research questions (Section1 below); variation from the diachronic
perspective, which takes the shape of tracing diachronic relations across linguistic patterns or studying change in specific linguistic categories (Section2); intra-lingual variation in terms of standard and non-standard varieties of a single language (Section3);
and variation we observe cross-linguistically, ranging from typological comparisons
to issues of language genesis and language death (Section4).
Variation spans truly all areas of language analysis: not just the traditionally accepted
structural and lexical-semantic domains but also, as this volume documents, all areas
of pragmatic categories and patterns. Interestingly, though, the modern history of diachronic and variationist discipline(s) reflects the fields journey through individual
domains, as research at different times has tended to zero in on different aspects of
linguistic structure: from the early interest in sounds, morphological structure, and word
etymologies as manifestations of speaker-independent systems to gradual inclusion of
syntactic patterning in texts to, most recently, incorporating patterns of language use
and the speakers role in language change. This shifting focus is determined in part by
the theoretical foundation one builds on and in part by the kind of empirical material
that is available at a given point in time, whether it is ancient texts documenting now
extinct languages, or searchable electronic corpora of authentic conversations and
other spoken discourse produced by present-day speakers.
Most of the contributions also argue some explicitly so, some more implicitly
for accepting and developing actively a systematic connection between research in
diachrony, synchronic variation, and typology. Integrating the methods and analyses
from these different perspectives can only lead to a better understanding both of specific historical data and of the general cognitive and communicative processes that
underlie the dynamic nature of language. The volume as a whole thus joins the voices of

Mirjam Fried

skepticism (cf. for example Nevalainen et al. 2006 or Croft 2006) regarding a categorical
division between diachrony and synchrony, which was instituted as one of the main
tenets in the structuralist thinking and has been maintained with even greater urgency
in generative linguistics. Such a division is inevitably called into serious question once
we acknowledge that synchronic language is inherently variable, which, in turn, can
be amply documented empirically once we turn to investigating linguistic patterns in
the context of actual discourse.
1. Approaches to variation and change
Questions of how languages change in time and what motivates the observed changes
frame the research concerns of historical linguistics, summarized in Goosens chapter.
While the primary empirical material that is traditionally associated with historical
linguistics comes from older stages of a given language, there is growing recognition
within general linguistics that the relevance of diachronic analysis goes beyond establishing genetic relationships between languages. It is becoming increasingly clear that
the contribution of diachronic studies reaches into other lines of research as well, such
as understanding the emergence of new grammatical patterns or lexical inventories
and their (re)organization, understanding language death or, conversely, the emergence of a new language (creolization), etc. Concrete research in historical linguistics
has reflected the theoretical background that informs the historical analysis and that
has changed over time, ranging from the textually grounded philological tradition to
the structuralist preoccupation with the effect of individual changes on the structure
and organization of the system as a whole, to the generativists emphasis on linking language change to child language acquisition, to the most recent theoretical approaches,
which are functionalist in nature and, to some extent, can be seen as a modern, more
sophisticated, cognitively grounded renewal of the philological tradition. In particular, the emergence of grammaticalization theory (to be discussed briefly in Section2),
prototype-based historical semantics (Geeraerts 1983, 1992, 1997; Luraghi 2003), and
the renewed appreciation of the value of old texts as valid sources of linguistic data
(e.g. Herring et al. 2000; Traugott & Dasher 2002) all have (re)established the relevance
of diachronic evidence and analysis to the general linguists quest for sources of explanation, which would improve our understanding of the cognitive and communicative
underpinnings of linguistic structure. This shift manifests itself in giving prominent
status to the textual dimension and greater attention to the pragmatic aspects of
linguistic change.
One way of practicing diachronic analysis is to compare discrete stages in the
history of a given language or a group of languages. Such a comparison can be set up
from two different perspectives: going forward in time, or going backward. The former

Introduction

is the primary concern of historical linguistics, which can thus be summarized as an

enterprise in tracing the history of a language from earlier (but, necessarily, attested)
stages to later stages of development. Whether the chronological point of departure
goes back many centuries or even millennia (e.g. to the first attestations of an extant
language, or in the study of dead languages), or just a handful of decades (e.g. in comparing a generational change in a particular linguistic feature), the method remains
the same: we are examining attested states of development and, perhaps, their relative,
or even absolute, chronology.
The latter perspective, i.e. going backward in time, is characteristic of diachronic
analyses in which the goal is to reconstruct an earlier, unattested stage, based on
general principles of language organization and on assumptions that can be made with
respect to expected mechanisms and pathways of change. However, reconstruction
as a method applies to two different sets of linguistic scenarios and analytic goals.
One is associated with the comparative method (discussed in detail by Nurse in this
volume), in which reconstruction is based on evaluating attested correspondences
between comparable features in multiple languages (or dialects) in order to establish
a hypothetical common ancestor (a proto-language). This kind of work has been
carried out particularly in the domain of phonetic and lexical inventories, on the basis
of which we can make hypotheses about proto-languages and about genetic relationships among languages.
The other type of reconstruction is known as internal reconstruction (briefly
addressed in Hickeys chapter), which is a procedure that aims at establishing a diachronic relationship between multiple variants within a single language, based purely
on inferring part of the history of a language from material available for a synchronic
description of the language, and from that alone (Chafe 1959; also Jeffers & Lehiste
1982). This method has been applied in various contexts, including accounts of synchronic polysemies or polyfuntional networks associated with grammatical items that
arise during grammaticalization processes (e.g. Traugott 1986). In that context, the
methods reliability can be greatly enhanced by grounding the reconstructive work in
a close analysis of the shifts in semantic-pragmatic patterns in actual communication
and thus draw on evidence from actual use (a recent example of such an application of
internal reconstruction can be found in Fried 2009).
Examining pragmatic motivations for semantic change is one of the areas of
research associated with the field of historical pragmatics, which is a relative newcomer in the field; its history, content, and methods are presented in Juckers contribution. Historical pragmatics brings to diachronic studies elements of discourse analysis
and an interest in issues that are generally associated with pragmatic principles, here
applied to diachronic material and goals: the evolution of politeness systems, speech
acts, discourse markers, and the like. Especially the latter two areas have benefitted
from the growing availability of sufficiently large and reliable electronic corpora and

Mirjam Fried

corpus-linguistic methods, without which the scope of diachronic pragmatic research

would, necessarily, remain fairly limited. Juckers chapter addresses the methodological
and data problems of diachronic pragmatics, but also sketches possible directions and
rich areas for additional research focused on various pragmatic aspects in the history
of specific languages, as well as on the need for comparative studies.
Viewed from the broadest perspective, language change may also be linked to the
question of language origin and evolution as a uniquely human behavior. This line of
research, discussed in Wildgens chapter, brings into diachronic analysis the concepts
of adaptation and selection, although they are treated differently depending on the
underlying linguistic theory. On the one hand, formally oriented research based on
the Chomskyan model of language reduces, rather unproductively, the issue of language evolution to the existence of the faculty of language in the narrow sense (Fitch
etal. 2005), which rests on the concept of recursion, taken to be the central operation
that defines language as a unique computational system. This approach necessarily
excludes all connection between language evolution, language change, and language
use. Socio-cognitive approaches, by contrast, make such a connection central in their
attempts to identify possible sources of language development and to trace evolutionary steps toward the emergence of full human language. Taking information-sharing
as a crucial concept, the socio-cognitive proposals rest on the study of the nature of
signs in animal and human communication, suggesting several basic prerequisites
for language evolution: categorization of the environment, the existence of a (structured) social group, and self-awareness. Functional explanations of the evolutionary
process then operate with various alternatives (possibly co-present and serving as
co-determinants) that may facilitate information sharing, such as the concept of reciprocity, relative size of the social group, and the use and manufacturing of tools. In fact,
a parallel is drawn between the pragmatics of tool use and the emergence of symbolic forms (from context-dependent, ad hoc tokens toward context-independent,
generalized patterns, which crucially involve planning and, thus, a particular kind of
abstraction). The study of language evolution is thus framed as a study of evolutionary
pragmatics, leading to the view that language emerges step by step from the interplay
of patterns of action and patterns of social structure in a (largish) group.
By establishing explicitly the contexts of use as the point of departure, evolutionary pragmatics defines its scope of research specifically as a branch of pragmatics and
emphasizes the social dimension of language evolution. However, the theory of evolution has made its way into general diachronic theorizing as well, in which researchers
grapple with the effects of application and, indeed, meaningfulness of evolutionary
concepts in explaining linguistic variation and change. A particularly detailed and
informative debate on the consequences of drawing direct parallels between biological
evolution and the processes that shape language change can be found in Thomsen
(2006), confronting two distinct views on the matter: Andersens (2006) approach, in

Introduction

which language change is argued to be inherently a cultural process (as would be

consistent with the approach of evolutionary pragmatics) and which necessarily resists
diachronic explanations framed by concepts relevant in biological evolution, contra
Crofts (2006) arguments, which are much more sympathetic to the explanatory potential
of viewing linguistic diachrony through the lens of processes and mechanisms known
from evolutionary biology. The debate will no doubt continue.
2. Diachronic analysis and pathways of change
Although variation and change can be seen as two facets of the same phenomenon a
synchronic variation always having the potential of being the intitial state of what may
become an observable diachronic change variationist research has tended to focus
either on identifying and describing patterns of synchronic variation, or on identifying
and describing diachronic change. Issues characteristic of the latter are taken up in
Hickeys comprehensive overview of the diachronic domain. By sketching the history
of the field and defining the constantly evolving research themes, scope of the data,
goals, methods, and theoretical relevance of studying language change, the chapter is,
in essence, a preview of many of the topics that are elaborated in the other chapters,
including the relevance of contact-based accounts, as well as areal and general typological issues vis--vis language change.
As Hickey also points out, the field of diachronic analysis started out as an enterprise in discovering and documenting instances of change and, especially in the
Neogrammarian and structuralist traditions, great emphasis was put on establishing
exceptionless regularity of change. However, in recent decades, diachronic research
has shifted its focus toward explaining instances of language change and suggesting
generalizations over those instances, thus laying the groundwork for identifying and
explaining recurrent types and direction of change. In the process, it has also become
increasingly clear that diachronic evidence bears on issues that are central to developing
an adequate theory of language in general and conversely, various models of language
have been brought to bear on possible explanations of variation and change.
One such model is discussed in Kearns chapter on implicature and language
change. She discusses various inferential patterns that may play a role in different types
of semantic change, appealing to Neo-Gricean theories of conversational principles,
and particularly to the Principle of Informativeness, as the relevant source of explanation. The chapter examines the properties of generalized conversational implicatures
as a special inferential mechanism, which can account for a semantic change that is
distinct from the products of either metaphor or metonymy. This kind of implicature
is posited as constituting the initial stage of a semantic change and is characterized
by inferences that lead to a more informative or more specific interpretation than

Mirjam Fried

the literal meaning offers (e.g. temporal > causal). In order to properly sort out the
relationship between this informativeness-strenghtening implicature and metonymic
processes (which in some accounts have been classified as closely related), the author
argues that we need a more nuanced analysis of sense relations associated with a given
form vis--vis the contexts in which the different senses can occur. Thus, metaphor
and metonymy both represent sense addition, resulting in a polysemy in which each
sense is associated with a different context, while conversational implicature, narrowly
defined as a process that strengthens informativeness, constitutes a pattern of meaning
substitution: both meanings are available simultaneously in a given context.
This type of change is also reminiscent of the concept of pragmatic strengthening (Traugott 1988), which has been invoked to account for a type of pragmaticsemantic change that is distinct from semantic bleaching and that may accompany
grammaticalization processes. In fact, the grammaticalization framework (e.g. Givn
1979; Traugott 1988, 1992, 1996, 2003; Hopper 1991; Heine et al. 1991; Bybee et al.
1994; Haspelmath 1998, 1999; Fischer et al. 2000; Traugott & Dasher 2002; Hopper&
Traugott 20032; Bisang et al. 2004; Van linden et al. 2010) is one prominent area of
research focused on establishing pathways of change, and one in which the role of context in linguistic change is given central importance. Grammaticalization research deals
with the emergence and evolution of grammar, originally conceived of as a process
in which independent lexical items with referential content become bound morphemes that serve the function of grammatical markers (cf. Meillet 1958[1912]:131).
Through the vigorous pursuit of grammaticalization phenomena over the past several
decades, it has become clear that the scope is much wider than the early interest in
identifying the source of inflectional morphemes, and the early definition has been
relaxed somewhat to apply to any process in which a morphosyntactically more complex and/or semantically more concrete expression develops into a formally simpler
and more abstract one, i.e. less grammatical (and not necessarily lexical) becomes
more grammatical.
From the beginning, the grammaticalization approach to language change has
combined two components: (i) the study of morphosyntactic change, originally motivated primarily by questions concerning the emergence of inflectional morphology
(cf. Lehmann 1995[1982]), but now extending to all types of grammatical entities;
and (ii) the study of semantic or functional change, originally accounting mainly for
the loss of referential meaning and concomitant decategorialization (noun > adposition, lexical verb > auxiliary, etc.) hence the notion of semantic bleaching. Most
recently, though, grammaticalization research especially in the domain of syntax has
been evolving toward including also changes that can be best labeled as constructionali
zation: processes that lead to the emergence of new syntactic patterns (grammatical
constructions) out of previously independent material, or to the reorganization of
existing ones. This perspective reflects two interrelated developments within the

Introduction

grammaticalization approach. One factor is the increasing awareness among its practitioners that grammatical change crucially involves recurring collocational patterns
and occurs gradually, through step-by-step adjustments in individual features that,
ultimately, affect the shape and grammatical status of the whole pattern; capturing
the true nature of change thus cannot be reduced to describing individual, contextfree grammatical items. The seeds of this line of thinking can be found already in
Lehmanns (1995[1982]: 406) explicitly worded formulation that grammaticalization
does not merely seize a word or morpheme [] but the whole construction formed
by the syntagmatic relations of the element in question, but it has been taken up more
systematically only fairly recently (for succinct explications of this idea, cf. for example
Traugott 2003a; Harris 2003; Wiemer 2004; Wiemer & Bisang 2004). However, this
development can be seen as a natural consequence of bringing pragmatics into the
study of meaning change that accompanies grammaticalization. The idea that the loss
of propositional meaning may be simultaneously accompanied by the emergence of
(inter)subjective meanings (e.g. Traugott 1982, 1989, 2003b) has opened up a line of
research in which language change is necessarily grounded in context-dependent and
speaker-dependent patterns of use.
These two factors are of course consistent with viewing grammar as an inventory
of form-meaning associations, such as we find in construction-grammatical approaches
to language. Some recent case studies that are explicitly oriented toward exploring the
details of a constructional basis of grammatical change and the importance of collocations as a vehicle in such changes can be found especially in Traugott 2007, 2008a,
2008b, or 2008c, among others (from the grammaticalization perspective as the focal
point) or Fried 2008, 2010 (from the construction-grammatical perspective as the focal
point), but also in other papers, e.g. in Leino 2008 or Bergs & Diewald 2008, 2009.
While there is continuing debate on what exactly should be accepted as the definitional criteria for grammaticalization (especially in contradistinction to lexicalization),
some consensus has been emerging lately and can be summarized in the following
list, which reflects Himmelmanns (2004:33) formulation (based directly on the conception developed originally in Traugotts work), elaborated by additional conditions
posited in Brinton & Traugott (2005: 108109): grammaticalization is a process of
conventionalization that crucially consists in the expansion of the semantic-pragmatic
usage contexts of the grammaticalized element and can involve (a subset of) additional
changes, including host-class expansion (i.e. the new behavior spreads from a specific
subset of items to all eligible members of a given class), the broadening of the syntactic
context in which a given item occurs, typological generality, and, potentially, subjectification. It remains a matter of further empirical research to establish whether these
changes form a hierarchy of sorts and if so, what affects their relative weight, which
may be typologically determined, or related to particular stages in the grammaticalization process, or dependent on yet other criteria. Another issue that continutes to be

Mirjam Fried

debated is the distinction between grammaticalization and lexicalization on the one

hand, and grammaticalization and pragmaticization on the other. The complexities of
the former are treated quite thoroughly in Himmelmann 2004 and Brinton & Traugott
2005 and the discussion in these works has moved us quite a long way toward sorting
out the relationship. The latter, however, concerning the change of an independent item
(whether lexical or grammatical) into a pragmatic marker, still awaits a comprehensive
treatment; so far it has been mostly subsumed under the notion of (inter)subjectification. Some recent case studies involving pragmaticization are presented in Fried
(2009) and in Gnthner & Mutz (2004), the latter also including a brief discussion of
the grammaticalization-pragmaticization relationship.
To summarize, the current orientation of grammaticalization studies conceives
of grammatical change as arising from multiple motivations and consisting of partial,
synchronically often imperceptible transitions that gradually (and collectively) lead
to an observable and internally more complex change. Moreover, since the transitions
are grounded in concrete instances of meaning negotiation in concrete communicative situations, the nature of the meaning change calls for a more sophisticated
analysis than the early emphasis on semantic weakening (or bleaching); hence the
idea of pragmatic strengthening and similar accounts based on the notion of informativeness. The objective of grammaticalization research is thus to search for explanations with respect to recurrent patterns of change and at the most general level,
those patterns are expected to be unidirectional. The unidirectionality hypothesis
follows from the observation that cross-linguistically, these processes overwhelmingly
involve a change that proceeds from lexical (or semantically more independent) to
grammatical (or semantically more abstract) status, rather than the opposite, in which
a grammatical item would acquire the status of an independent lexical unit with
propositional semantics.
3. Intra-lingual variation
As already noted, the study of linguistic variation tends to focus on the synchronic
relationships between variant patterns and it takes as its focal point either a more
local perspective (within a single language), or a more global one (across languages).
The distinction is more a matter of degree and emphasis than a matter of substance
or fundamentally different goals, but it has been quite persistently maintained, until
fairly recently. Regardless of this division, though, the present volume provides ample
evidence that the synchronically oriented variationist research shares an important
feature with the trends in diachronic research: namely, there is a noticeable shift from
merely describing concrete tokens of variation toward seeking generalizations.
This section is an overview of the local perspective, which consists in the study of
closely related synchronic varieties of one language. Its evolution as a field of research,

Introduction

traditionally known as dialectology, is the topic of De Schutters chapter. Perhaps the

most conspicuous development within the field concerns the broadening of its scope
in several, partially interrelated, directions. Firstly, the object of inquiry in dialectology
now includes language varieties that are not identified purely geographically but are
based also on social stratification and various socially based groupings that are linguistically sufficiently distinct in relation to each other (in terms of ethnicity, socioeconomic status, age, gender, etc.) and to some presupposed standard language. It is
worth mentioning that especially the so-called town dialects, i.e. varieties spoken in
regional centers, have been holding the dialectologists sustained interest in the most
recent decades. Secondly, this expanded scope brings in an explicit connection to sociolinguistic dimensions of dialectal research, incorporating topics such as politeness,
speech acts, turn-taking, topic-selection and management, discourse markers, etc.,
none of which ever entered dialectological research in the traditional sense. And
thirdly, the domain of inquiry in traditional dialectology is restricted primarily to
variation in sound patterns, lexical choices, and morphological structure, with only
minimal interest in syntactic variation. However, syntax is also becoming an active
and lively area of research, known as dialect syntax and pursued as a distinct subdomain, whose goal is to integrate the contributions of dialectological research on
syntactic patterning with syntactic theory.
A concrete conception of the pragmatically and sociolinguistically based approach
to dialectal variation is articulated in Schneiders chapter, which presents the research
as a specific branch of empirical pragmatics, namely, variational pragmatics. This
intersection of pragmatics and dialectology enriches both sides: it adds the pragmatic
dimension to the analysis of dialects, including conversation analysis, discourse analysis,
and interactional linguistics, but it also provides a conceptual space for studying variation across specific pragmatic categories. Moreover, a parallel can be drawn between
the rise of variational pragmatics and the rise of dialect syntax in terms of their larger
objectives and potential contributions to language analysis in general. Both are relative newcomers in the field of variationist research and both are concerned with
going beyond just identifying and describing the intra-lingual differences within their
respective areas; one of the research objectives in both is to search for and test appropriate explanatory models, be it in a particular syntactic theory (as practiced in certain
strands of dialect syntax), or in the socio-cognitive domain (through the concept of
cultural models, constructional conceptions of grammatical structure, etc.).
The fact that dialectology as a field has been with us since the Neogrammarians
does not mean that there is necessarily a firm consensus about the exact content
of the fields central notion and its terminological label, be it the more traditional
dialect or its more recent alternative, variety (among other terminological candidates).
Macaulays chapter aims at sorting out the difficulties in defining the concept of dialect,
through careful explication of the various terminological alternatives, but he comes
to the conclusion that the exact definition remains elusive. This is partly due to the

10

Mirjam Fried

increasingly richer inventory of linguistic entities to which it can be applied (as laid
out by De Schutter), but, as Macaulay emphasizes, the difficulties partly follow from
the very nature of linguistic variation: the speech differences that collectively amount
to a noticeably distinct variety resist straightforward categorization with clear-cut
boundaries between two adjacent varieties, whether we understand adjacent in the
geographical (horizontal) sense or as contiguity across social stratification (vertical). Put
differently, it is essentially impossible to state categorically where one variety ends and
another one begins in terms of the set of features that might completely and uniquely
differentiate one from the other.
Nevertheless, detailed and increasingly more sophisticated research on individual
varieties continues, and we get a glimpse of the relevant issues in this volume as well.
It may be hard enough to develop a general consensus concerning our understanding
of linguistically determined dialect boundaries, as an issue of establishing distinct
speech communities. Matters become even more complicated when we move to the
level of situation-related speech differences (registers), which may or may not intersect with dialect differences. The research questions concerning situational stratification of speech varieties is taken up in Dittmars chapter. Finally, Martn Rojos chapter
on jargon is instructive in presenting the range of challenges dialect research may
face. It is not only the complexity and elusiveness of the subject matter, but as she
documents on the history of the studies of jargon linguistic research itself, as a
scholarly pursuit, may be affected (and distorted) by social attitudes. Those determine not only the patterns of speakers evaluation, acceptance, or rejection of specific
linguistic features but have also played a significant role in the status of jargon as an
object of scientific inquiry.
On the whole, we observe a double shift in dialect-related research: in the object of
study (from describing individual varieties as whole language systems toward describing
and incorporating features of dialect use) and in the method and general goals (from
mainly comparative research whose primary objective is to provide evidence relevant
to reconstruction toward examining the status of linguistic variation in actual communication). As emphasized particularly in Schneiders chapter, this does not mean
that the comparative dimension should be abandoned on the contrary; it simply
means that the comparative work needs to be extended to categories in the pragmatic
domain, which had not played a role in dialectal research until very recently.

4. Cross-language variation
The global perspective of variation is, by definition, the central concern of typological
studies. In typology, the aim is to discover constraints on variation across languages
and principles that account for the observed variations, in order to establish general

Introduction

(and possibly universal) properties of human languages as well as the range of potential
differences among languages. In the present volume, typological issues are addressed
primarily in Comries chapter although some large-scale typological changes are also
touched upon by Hickey. While typological research continues to make major contributions to our understanding of the patterns of variation in grammatical systems, it
is only beginning to address the pragmatic dimension of linguistic organization. Yet,
the typologists emphasis on a solid empirical footing, which is seen as a prerequisite
for articulating meaningful generalizations, makes a natural connection to pragmatics,
which, in turn, provides potential models that typologists can explore in their search
for general explanatory principles. Thus, a systematic study of language use, either
concerning individual languages or in cross-linguistic comparisons (whether framed
by contact situations, or without contact), charts a new direction in which we can
investigate the socio-cognitive grounding of grammatical systems and propose empirically supported typologies that will include pragmatic variation.
An area in which cross-linguistic variation has been approached primarily from
the pragmatic and psycholinguistic perspective, so far without much concern for any
potential typological consequences, concerns situations in which two or more languages are in intense close contact; a crystallized form of such a contact is a bilingual
(or multilingual) situation, manifested either in a bi-/multilingual individual or in a
bi-/multilingual society (in which not all speakers are necessarily bi-/multilingual).
An investigation of language contact as a source of change, or motivation for change,
necessarily emphasizes the pragmatic dimension of linguistic analysis, since the potential consequences of language contact cannot be reduced to cataloguing particular
changes in the linguistic structure or vocabulary of the contact language systems; contact may give rise to changes both in linguistic structure and in patterns of usage. As
Matras chapter makes clear, a crucial ingredient in contact-induced changes is their
grounding in the specifics of the discourse context: the topic, the status of the interlocutors, the situation, conversational agendas, etc. (echoed also in Meewis & stman
in this volume). Incorporation of these factors, together with the psycholinguistic and
cognitive perspective taken in many studies, is not only consistent with the search for
explanatory models that might prove helpful in explaining contact-induced changes,
but also moves us away from the simplistic view that mere reference to contact constitutes an explanation. Moreover, research on language contact clearly documents and
explicitly acknowledges that linguistic evidence from contact situations bears on many
general questions concerning the relation between language and the brain/mind, such
as language acquisition in bilingual (or multilingual) children; language processing
issues, such as storage, memory, and activation; cognitive factors that determine patterns
of code-switching; etc.
As the research involving contact phenomena and contact situations has gained
a stable place not only within pragmatics but also in connection to much broader,

11

12

Mirjam Fried

theoretical questions about linguistic organization, it is now possible to gather the

associated research agendas under the heading of contact linguistics, introduced in the
chapter by Meeuwis & stman as a distinct domain related to the study of language
change. A crucial premise and a fundamental contribution of all contact-related
research is the explicitly stated acknowledgment that contact happens between speakers,
not between languages (as pointed out also by Hickey). Consequently, linguists attention must be directed at the social factors that play a role in contact-induced developments, such as the social status of discourse participants, the nature and intensity of
cultural contact, etc. Moreover, once we take a systematic approach to studying contact
phenomena, we can gradually move beyond just describing single features and toward
identifying recurrent patterns in contact-induced changes and toward articulating
more general conclusions about what exactly is or is not likely to be affected by language
contact: e.g. the relative effect on open-class items vs. closed-class items; lexicon vs.
grammar; whole language vs. only parts of the system or individual features. With
respect to the latter contrast, a well-known example of contact phenomena that involve
only one part of a linguistic system is borrowing, traditionally confined to matters of
lexical borrowing. This type of borrowing was among the earliest domains of scholarly
interest in contact-motivated changes although its original conception was even more
narrowly oriented, mostly reducing borrowing to questions of cultural loans.
While the phenomenon of borrowing has always had its place in the studies of
language change, especially as one of the tools that can aid in establishing genetic relationships between languages, Treffers-Dallers chapter shows that the field has become
richer and considerably more sophisticated over time. It has clearly expanded its interests
beyond the traditional focus on lexical borrowing, not only toward investigating other
types of borrowing, including grammatical categories or entire grammatical patterns
as well as pragmatic loans, but also, most recently, toward posing larger questions that
aim at identifying generally applicable principles of borrowability. All of this means
paying attention to constraints on borrowing and treating borrowing processes in a
more systematic manner. Especially the incorporation of the pragmatic dimensions
of borrowing has drawn attention to the role of speakers in the borrowing process,
which brings forward also issues of actuation and spread a quintessential preoccupation of historical linguists.
Perhaps a related problem but one that arises in a very specialized contact situation
is the question of transferrability of linguistic features, including transfers of pragmatic categories and patterns, in the context of second language acquisition. The field
of interlanguage pragmatics, introduced in Kaspers contribution, is a subfield of pragmatics, whose research agenda is framed by the central concept of interlanguage: the
transitional (and transient) form of the target language that second language learners
develop during the acquisition process. As the label indicates, the primary focus is
on the pragmatic dimension of this process, i.e. the transferability of conversational
patterns and strategies. The research is necessarily comparative in nature, since the

Introduction

point is to investigate the relation of the interlanguage to both L1 and L2 and to capture the scope (and limits) of variation within the interlanguage. In this enterprise,
variation is investigated from the perspective of transferability, in recent studies turning attention also to more general questions concerning the conditions for pragmatic
transfer and the interaction between various factors involved in the transfer. It must
also be noted, though, that research of this kind has to cope with the fact that assessing the relative success of the L2 learner has to take into consideration the inherent
variation in L1 as well; it is not always clear what the standard of comparison (i.e. a
norm) is or ought to be, against which the state of the interlanguage can be measured. For now, interlanguage research has been carried out mostly in a synchronic
mode, biased toward studying instances of language use, rather than tracing development of pragmatic competence over time. However, the potential of contributing to
diachronic analyses as well is evident and can constitute an additional direction in
interlanguage research.
Finally, Mufwenes chapter addresses another highly specialized type of contactmotivated development creolization, i.e. the genesis of a new vernacular in a particular kind of contact situation and under particular socio-pragmatic conditions.
Questions that are internal to creole studies and at the same time relevant to variationist concerns include primarily the shape of the structural design (is it the same
for all creoles, or do they differ from each other?); the socio-pragmatic conditions for
their development (again, same or different?); and the acquisition mechanism (is their
development crucially dependent on child language acquisition, or could creolization
processes take place in adult speakers as well, especially considering certain similarities
between creolization and second-language acquisition?). Other problems, however,
reflect more general challenges, common to all dialectal research. In particular, we
observe in creole research a direct link to the difficulties associated with identifying
and defining dialects, in contradistinction to separate languages: in the absence of clear
and consistent structural criteria for defining creoles as distinct from non-standard
varieties of their lexifiers, it may not be wholly surprising that there is room for the
known and apparently persistent controversy involved in classifying non-standard
varieties of various European languages, spoken by non-European communities.
However, with growing availability of electronic corpora, which provide easier access
to the relevant material, there is hope that a more detailed and precise grammatical
analysis will be carried out, providing evidence that could serve as a basis for a more
refined and consistent comparison between the varieties in question. An interesting
corpus-based study that points in this direction, documenting structural similarities
between spoken Parisian French and varieties spoken by non-French speakers outside
of France can be found in Gadet (In press). It is clear that research on creoles, their
genesis and their structural properties, offers great potential for contributing to general
linguistic theorizing (not just in the acquisition domain, but also in diachronic studies,
and especially grammaticalization) as well as to typology.

13

14

Mirjam Fried

5. Summary and outlook

The inherently dynamic nature of language manifests itself in many ways, each of
which continues to inspire focused research with emphasis on particular sets of issues,
ranging from the comparative reconstruction of ancestral languages a dominant
preoccupation in the early days of historical linguistics to establishing the properties
of interlanguage (the transient state of linguistic competence in second language
acquisition), which is among the more recent additions to the ever widening discipline.
To some degree, the present volume captures the evolution of the field itself, which
proceeds along a number of dimensions: (i) from attention to morpho-phonological
and lexical systems toward analyzing syntactic and pragmatic patterning; (ii) from
comparative reconstruction and text-oriented philology to the present-day functionalist
and socio-cognitively oriented accounts; (iii) from the speaker-free approach in the
Neogrammarian, structuralist, and generativist analyses toward developing the sociolinguistic dimension of variation as a sui generis area of research, oriented toward
the interactional grounding of linguistic structure; and (iv) from intra-lingual scope
of variation toward a systematic inclusion of intercultural communication and thus
richer cross-linguistic research.
We thus register a qualitative shift, clearly consistent with the developments of the
last forty or so years within general linguistics, with its interest in explanations and
the prominent role of linguistic theory: variationist and diachronic research continues
to move beyond cataloguing instances of variation and toward developing theoretically informed accounts that aim at explaining variation and change. Some subfields
have moved further along in this effort than others but the awareness that plausible
explanations are the ultimate goal is evident in all of them. The contributions in the
present volume argue for incorporating the socio-cognitive perspective in linguistic
analysis as a particularly promising source of useful methodology and explanatory
models and provide concrete examples of implementing the dictum that, in Crofts
(2006:107) words, [i]n order to understand the mechanisms of language change, one
must examine social interactions and human relationships, not just linguistic forms
and their frequencies.

References
Andersen, H. (2006). Synchrony, diachrony, and evolution. In N. Thomsen (ed.), 5990.
Bergs, A. & G. Diewald (2008). (eds.) Constructions and language change. Mouton de Gruyter.
(2009). (eds.). Contexts and constructions. John Benjamins.
Bisang, W., N.P. Himmelmann & B. Wiemer (2004). (eds.). What makes grammaticalization? A look
from its fringes and its components. Mouton de Gruyter.

Introduction

Brinton, L.J. & E. Closs Traugott (2005). Lexicalization and language change. Cambridge University
Press.
Bybee, J., R. Perkins & W. Pagliuca (1994). The evolution of grammar: tense, aspect, and modality in
the languages of the world. Chicago University Press.
Chafe, W. (1959). Internal reconstruction in Seneca. Language 35: 477495.
Croft, W. (2006). The relevance of an evolutionary model to historical linguistics. In N. Thomsen (ed.),
91132.
Fischer, O., A. Rosenbach & D. Stein (2000). (eds.) Pathways of change: grammaticalization in English.
John Benjamins.
Fitch, W.T., M.D. Hauser & N. Chomsky (2005). The evolution of the language faculty: clarifications
and implications. Cognition 97: 179210.
Fried, M. (2008). Constructions and constructs: mapping a shift between predication and attribution.
In A. Bergs & G. Diewald (eds.), 4779.
(2009). Construction Grammar as a tool for diachronic analysis. Constructions and Frames
1/2: 261290.
(2010). Grammaticalization and lexicalization effects in participial morphology: a Construction
Grammar approach to language change. In A. Van linden, J-Ch. Verstraete & K.Davidse (eds.).
Gadet, F. (In press). What can be learned about the grammar of French from corpora of French
spoken outside France. In M. Konopka, J. Kubczak & U.H. Waner (eds.), Proceedings of the
Third International Conference Grammar and Corpora. Gunter Narr Verlag.
Geeraerts, D. (1983). Reclassifying semantic change. Quaderni di Semantica 4: 217240.
(1992). Prototypicality effect in diachronic semantics: a round-up. In G. Kellermann &
M.D.Morrissey (eds.), Diachrony within synchrony: language history and cognition, 183204.
Peter Lang.
(1997). Diachronic prototype semantics: a contribution to historical lexicology. Clarendon.
Givn, T. (1979). On understanding grammar. Academic Press.
Gnthner, S. & K. Mutz (2004). Grammaticalization vs. pragmaticalization? The development of
pragmatic markers in German and Italian. In W. Bisang et al., 77107.
Harris, A.C. (2003). Cross-linguistic perspectives on syntactic change. In B.D. Joseph & R.D. Janda
(eds.), The handbook of historical linguistics, 529551. Blackwell.
Haspelmath, M. (1998). Does grammaticalization need reanalysis? Studies in Language 22: 315351.
(1999). Why is grammaticalization irreversible? Linguistics 37: 10431068.
Heine, B., U. Claudi & F. Hnnemeyer (1991). Grammaticalization: a conceptual framework. Chicago
University Press.
Herring, S.C., P. van Reenen & L. Schsler (2000). (eds.) Textual parameters in older languages.
John Benjamins.
Himmelmann, N.P. (2004). Lexicalization and grammaticalization: opposite or orthogonal? In
W.Bisang, N. Himmelmann & B. Wiemer (eds.), 1940.
Hopper, P. (1991). On some principles of grammaticalization. In E. Closs Traugott & B. Heine (eds.),
Approaches to grammaticalization I, 1735.John Benjamins.
Hopper, P. & E. Closs Traugott (20032). Grammaticalization. Cambridge University Press.
Jeffers, R.J. & I. Lehiste (1982). Principles of historical linguistics. The MIT Press.
Lehmann, Ch. (1995 [1982]). Thoughts on grammaticalization. LINCOM Europa.
Leino, J. (2008). (ed.). Constructional reorganization. John Benjamins.
Luraghi, S. (2003). On the meaning of prepositions and cases: the expression of semantic roles in Ancient
Greek. John Benjamins.

15

16

Mirjam Fried
Meillet, A. (1958 [1912]). Lvolution des formes grammaticales, reprinted in Linguistique historique
a linguistique gnrale, 131148. Champion.
Nevalainen, T., J. Klemola & M. Laitinen (2006). (eds.). Types of variation: diachronic, dialectal and
typological interfaces. John Benjamins.
Thomsen, N. (2006). (ed.). Competing models of linguistic change: evolution and beyond. John
Benjamins.
Traugott, E. Closs (1982). From propositional to textual and expressive meanings: some semanticpragmatic aspects of grammaticalization. In W.P. Lehmann & Y. Malkiel (eds.), Perspectives on
historical linguistics, 245271. John Benjamins.
(1986). From polysemy to internal reconstruction. BLS 12: 539550.
(1988). Pragmatic strengthening and grammaticalization. BLS 14: 406416.
(1989). On the rise of epistemic meanings in English: an example of subjectification in semantic
change. Language 65/1: 3155.
(1992). From less to more situated in language: The unidirectionality of semantic change.
In M. Kefer & J. van der Auwera (eds.), Meaning and grammar: cross-linguistic perspectives,
497517. Mouton de Gruyter.
(1996). Grammaticalization and lexicalization. In K. Brown & J. Miller (eds.), Concise encyclopedia of syntactic theories. Elsevier.
(2003a). Constructions in grammaticalization. In B.D. Joseph & R.D. Janda (eds.), The
handbook of historical linguistics, 624647. Blackwell.
(2003b). From subjectification to intersubjectification. In R. Hickey (ed.), Motives for language
change, 124139. Cambridge University Press.
(2007). The concepts of constructional mismatch and type-shifting from the perspective of
grammaticalization. Cognitive Linguistics 18/4: 523557.
(2008a). Grammaticalization, constructions and the incremental development of language:
Suggestions from the development of degree modifiers in English. In R. Eckardt, G. Jger &
T.Veenstra (eds.), Variation, selection, development probing the evolutionary model of language
change, 219250. Mouton de Gruyter.
(2008b). The grammaticalization of NP of NP patterns. In A. Bergs & G. Diewald (eds.), 2345.
(2008c). All that he endeavoured to prove was: On the emergence of grammatical constructions in dialogic contexts. In R. Cooper & R. Kempson (eds.), Language in flux: dialogue
coordination, language variation, change and evolution, 143177. Kings College Publications.
Traugott, E. Closs & R.B. Dasher (2002). Regularity in semantic change. Cambridge University Press.
Van Linden, A., J.-Ch. Verstraete & K. Davidse in collaboration with H. Cuyckens (2010). (eds.), Formal
evidence in grammaticalization research. John Benjamins.
Wiemer, B. (2004). The evolution of passives as grammatical constructions in Northern Slavic and
Baltic languages. In W. Bisang et al., 271331.
Wiemer, B. & W. Bisang (2004). What makes grammaticalization? An appraisal of its components
and its fringes. In W. Bisang et al., 320.

Borrowing
Jeanine Treffers-Daller

1. Introduction
Borrowing the incorporation of features of one language into another has been
studied by researchers working in a wide range of areas, from a diachronic as well as
a synchronic perspective. In a diachronic research tradition that focuses on the historical development and the genetic classification of languages, it is clearly of central
importance to be able to distinguish borrowed features from non-borrowed or native
features of a language. The importance of the role of lexical and structural borrowing
for language change has however only fairly recently been recognised, as until the
publication of Thomason and Kaufmans (1988) book on contact-induced change,
historical linguists used to emphasise internal causes of language change. Interest in
the synchronic analysis of borrowing emerged towards the turn of the century. The
main focus of the synchronic analyses has been to identify the grammatical constraints
on borrowing, and to describe the phonological, syntactic and morphological integration of borrowed words. In addition, researchers have tried to delimit borrowing from
other language contact phenomena, such as code-switching and transfer, and they
have developed different classifications of borrowing. The social correlates of borrowing have received attention in more quantitatively oriented studies. Put differently,
researchers have mainly focused on what Weinreich, Herzog and Labov (1968) have
called the embedding problem and the constraints problem. The embedding problem,
when applied to the study of borrowing, concerns on the one hand the embedding or
integration of source language features in the borrowing language. On the other hand,
it deals with the embedding of these features in the social structure: to what extent do
social factors influence the quantity and the quality of the borrowing process. The constraints problem deals with the question of determining the set of possible borrowings
and with the discovery of the structural constraints on the borrowing process. Other
questions have received less systematic attention. The actuation problem and the transition problem (how and when do borrowed features enter the borrowing language
and how do they spread through the system and among different groups of borrowing language speakers) have only recently been studied. The evaluation problem (the
subjective evaluation of borrowing by different speaker groups) has not been investigated in much detail, even though many researchers report that borrowing is evaluated

18

Jeanine Treffers-Daller

negatively. Apart from the issues raised above, in more recent studies, pragmatic and
psycholinguistic aspects of borrowing have been studied in some detail.
2. Short historical overview
Early synchronic studies on borrowing, such as Salverda de Grave (1906), focus on
lexical borrowing, and often consist of word lists of borrowed words, the history of
each of which is discussed in some detail. The focus is generally on one language pair
and on borrowing in one direction only. In some studies the loan words are grouped
into different semantic categories and there is often some attention for the adaptation
of these words to the phonological system of the borrowing language. The syntactic
and morphological integration of borrowed words, on the other hand, remains often
unexplored, and pragmatic or psycholinguistic considerations are entirely absent.
The importance of these early studies resides in the meticulous detail with which the
history of individual words is analysed and in the fact that they show how important
lexical borrowing can be in a particular variety. These studies are limited in scope,
because it is not possible to predict with any confidence that a word cannot be borrowed
(Romaine 1995:142). Although most researchers no longer attempt to provide complete lists of borrowed words in a particular variety, an important new resource for
those interested in lexical borrowing can be found in the World Loanword Database (Haspelmath & Tadmor 2009), which provides an overview of borrowings in
41languages (http://wold.livingsources.org/).
It is mainly through the work of Bloomfield (1933), Haugen (1950) and Weinreich
(1953) that language contact was established as a new research field in the course of
the twentieth century. At first, the focus was mainly on lexical borrowing. This may
well be due to the fact that lexical borrowing is frequent in the languages of the world.
In addition, Weinreich (1953:56) points to the fact that words are more likely to be
borrowed than structural elements, and states that the vocabulary of a language,
considerably more loosely structured than its phonemics or its grammar, is beyond
question the domain of borrowing par excellence.
After the publication of the seminal works of Haugen and Bloomfield, the focus
of research shifted from the result of borrowing to the process of borrowing and the
principles behind this process. Many researchers discussed the grammatical principles constraining the occurrence of lexical borrowing. Some authors formulated
these general principles in the form of implicational scales which aimed at establishing universals of borrowing (Moravcsik 1978) and others formulated hierarchies of
borrowability (Muysken 1981a). The first large quantitative studies, which appeared
towards the end of the eighties (Poplack, Sankoff & Miller 1988), made it possible to

Borrowing

empirically test these hierarchies in large databases, and to investigate the social
correlates of the borrowing process in some depth. Furthermore, after the pioneering
studies of Pfaff (1979) and Poplack (1980), a large number of studies focused on codeswitching, the alternate use of two or more languages in discourse. Many researchers
concentrated upon finding criteria to delimit borrowing from code-switching, because
they felt that the search for constraints on code-switching was seriously impeded by
the problems involved in distinguishing code-switching from borrowing (see Appel &
Muysken 1987; Romaine 1995 for an overview).
Researchers agree that lexical borrowing is widespread in the languages of the
world, but the existence of structural borrowing has been the subject of many debates.
Thomason and Kaufmans (1988) work represents a turning point in language contact
research, as it is the first study which offers a comprehensive framework for the analysis
of a wide range of language contact phenomena, such as borrowing, interference and
convergence, and of contact languages, such as pidgins, creoles and mixed languages
(see below for more details). Also, they are the first to provide detailed evidence of the
role of external factors (language contact) in language change and they provide examples
and analyses of contact-induced change from a wide range of language pairs.
Prince (1988) is one of the first studies in which pragmatic borrowing is explored.
In the early nineties, pragmatic aspects of and constraints on borrowing are beginning to
be investigated, especially in those studies which focus on discourse markers (Salmons
1991; De Rooij 1996 & Matras 1998). In addition, important new insights are gained
from psycholinguistic research into the processing of bilingual speech (see Grosjean
1997 for an overview). Another new research strand concentrates on the analysis
of bilingual intonation patterns. Queen (2001:56) provides initial evidence that the
intonation patterns of her Turkish-German bilingual informants differ from those
of monolinguals, but according to Queen the result is a pattern that occurs in the
speakers Turkish and their German, yet is common to neither language. The mechanism behind this phenomenon is therefore not considered to be borrowing or transfer
but fusion. Colantoni and Gurlekistan (2004), on the other hand, point to a combination of direct and indirect transfer from Italian as the source of intonation patterns
found in Spanish as spoken in Buenos Aires.
3. Definitions of borrowing: Terminological issues
Defining borrowing is problematic because researchers use different terminology to
refer to a number of different language contact phenomena and it is difficult to find
generally accepted terminology. The definition given by Thomason and Kaufman
(1988) in their pioneering work on contact-induced change has been very influential

19

20

Jeanine Treffers-Daller

in the past decade and is therefore probably a good starting point. Thomason and
Kaufman (1988:37) define borrowing as follows:
Borrowing is the incorporation of foreign features into a groups native language by
speakers of that language: the native language is maintained but is changed by the
addition of the incorporated features.

In most cases, in a situation of language contact between bilinguals, words are

the first elements to enter a borrowing language. According to Thomason and
Kaufman, if bilingualism is wide-spread and if there is strong long-term cultural
pressure from source-language speakers on borrowing language speakers, structural
features may be borrowed as well phonological, phonetic and syntactic elements
and even (though more rarely) features of the inflectional morphology (Thomason &
Kaufman 1988:37).
Thomason and Kaufman hold the view that as far as the strictly linguistic possibilities go, any linguistic feature can be transferred from any language to any other language
(T & K 1988:14). It is clear, however, that not all features are equally likely to be borrowed. Thomason and Kaufman therefore propose a very detailed borrowing hierarchy,
which has subsequently been used widely by other researchers as a yardstick against
which the depth of the borrowing process in contact situation can be measured. The
hierarchy is a detailed elaboration of the hierarchies of borrowability that have been
proposed earlier in the literature (see below under constraints for more details).
The terminology adopted by Thomason and Kaufman is somewhat different from
the terminology used by other researchers. Haugen (1950:212) defines borrowing as
the attempted reproduction in one language of patterns previously found in another.
The problem of this definition is that the notion patterns remains rather vague, and
it is unclear to what extent elements beyond the word level are included. Weinreich
(1953:1), on the other hand, uses the term interference as a cover term for interlingual influence at different levels (phonological, syntactic, semantic and lexical) and
defines interference as those instances of deviation from the norms of either language
which occur in the speech of bilinguals as a result of their familiarity with more than
one language.
Van Coetsems definition of borrowing as recipient language agentivity corres
ponds to Thomason and Kaufmans definition of borrowing, and Van Coetsems (1988)
concept of imposition (source language agentivity) is relatively close to Thomason and
Kaufmans concept of interference through shift. Because of the existing confusion
over terminology, some researchers have proposed new terms. Clyne (1967 and 2003)
uses the term transference, partly because of the negative connotations attached to the
notion interference. Johanson (1993) introduces the term code copying, as the term
borrowing suggests that materials borrowed at some stage will be given back at a later
point, which is not the case. Milroy (1997:311) points out that as for language contact,

Borrowing

it is not actually languages that are in contact, but the speakers of the languages. I agree
with Milroy that there is a danger in seeing languages as discrete entities independent
of speakers, because under this view the role of speakers in actuating and diffusing
language change (or borrowing) is being neglected. Most researchers have however
continued to use the term borrowing, albeit in slightly differing definitions.
Researchers have different views on the necessity to distinguish borrowing from
other language contact phenomena, such as transfer, convergence and code-switching.
Poplack and associates (Poplack 1980; Poplack & Meechan 1995) maintain that borrowing and code-switching are different phenomena. Support for this position comes
from Grosjean and associates (Grosjean 1988, 1995, 1997), who have extensively
studied psycholinguistic aspects of code-switching and borrowing (see under psycho
linguistic approaches for more details). Other researchers assume that there is a
common set of formal principles to morphological and syntactic structure and that
as a result, there may well be parallel constraints on borrowing and code-switching
(Appel& Muysken 1987; Muysken 1990). The issue cannot be discussed here in more
detail (cf. Treffers-Daller 2005, 2009).

4. Different types/classifications of borrowing

Bloomfield (1933) is one of the first studies in which an attempt is made at classifying
lexical borrowing. He distinguishes between dialect borrowing, where the borrowed
features come from within the same speech-area (as, father, rather with [a] in an
[]-dialect), and cultural borrowing, where the borrowed features come from a different
language. (Bloomfield 1933:444). The term cultural borrowing is reserved for the
importation of words for cultural novelties, such as spaghetti from Italian, which are
introduced to the culture of the borrowing language. Clearly this is an increasingly
important phenomenon, especially in relation to the influence of (American) English
on other languages. Cultural borrowing is not necessarily one-sided, and intensive
contact between speakers of both languages is not a prerequisite. When speakers of
different languages come into more intensive contact, borrowing extends to speechforms that are not connected with cultural novelties (Bloomfield 1933: 461). This
is called intimate borrowing and it is generally one-sided: borrowing goes predominantly from the upper language to the lower language, that is from the culturally,
politically or economically dominant language speakers to the speakers of the less
prestigious language.
Haugen (1950, 1953) and Weinreich (1953) further develop the typology of lexical
borrowing. Haugens approach is new in that he also discusses the structural constraints on borrowing and the structural effects on the borrowing language as a
whole. Weinreich not only discusses lexical borrowing (or in his terminology: lexical

21

22

Jeanine Treffers-Daller

interference), but also syntactic and phonological interference. Weinreich also establishes links with the sociolinguistic and psycholinguistic theories of his time, and this
distinguishes his work from all other approaches to borrowing developed before. The
sociolinguistic and psycholinguistic aspects of his work are now out of date, but
the classifications of different types of borrowing Weinreich and Haugen developed
are still widely used. As they are very similar, the following discussion is limited to
Haugens typology.
Haugen (1950) distinguishes different types of borrowing, based on the question
of whether or not source language morphemes are imported into the borrowing and
whether or not substitution of source language morphemes or phonemes by borrowing
language elements occurs (see also Backus & Dorleijn 2009 for a further refinement
of this typology). We will illustrate the concepts with examples from French-Dutch
language contact data, as described in Treffers-Daller (1994).

a.Loanwords: Morphemic importation without substitution

When French-Dutch bilinguals import the French discourse marker donc (so) into
Brussels Dutch, they keep the French nasal vowel [f] and do not replace it with one or
more Dutch phonemes.

b.Loan blends: Morphemic importation as well as substitution

Bilingual speakers in Brussels use pertang (however), originally from French pourtant.
Because the French nasal vowel [c] has been replaced by the vowel [a] and the nasal
velar [] and the vowel [u] has been substituted by a schwa, this word is often not recognised as a borrowing anymore. The category of loan blends includes hybrids or mixed
compounds, which consist of French and Dutch morphemes, such as gemeente-taxe
(council tax), where gemeente (council) is Dutch and taxe (tax) is Brussels French.

c.Loanshifts: Morphemic substitution without importation

In the case of loan shifts only a meaning, simple or composite, is imported, but the
forms representing this meaning are native (Appel & Muysken 1987: 165). Famous
examples are German Wolkenkratzer, French gratte-ciel and Spanish rascacielos, all of
which are modelled on English skyscraper (Haugen 1950:214). In these cases both
halves of the compound have been translated into the borrowing language, but no source

Borrowing

language morphemes have been imported into the borrowing language. Similarly, in
some cases, the meaning of a word can be extended or changed without any importation of lexical material. An example of this phenomenon, called semantic loan, is found
in the extension of the meaning of Dutch tellen (to count). In Belgian Dutch, tellen has
acquired the meaning to count on from French compter sur, as in the expression:
iemand waar ge op kunt tellen (someone you can count on).
Poplack and associates introduced a different typology of borrowing, based on the
diffusion of these elements throughout a speech community:
Established loanwords (which typically show full linguistic integration, nativelanguage synonym displacement and widespread diffusion, even among recipientlanguage monolinguals) differ from nonce borrowings only insofar as the latter need
not satisfy the diffusion requirement (Poplack & Meechan 1995:200).

Grosjean (1995:262) refers to nonce borrowings as idiosyncratic loans or as speech

borrowings, and Appel and Muysken (1987) use the term lexical interference for the
same phenomenon.
The fact that borrowing can also take place between spoken and sign languages has
only recently started to be explored. Sandler and Lillo-Martin (2006:95) use the term
cross-modal borrowing for two kinds of borrowing in sign languages: the mouthing
of spoken words and fingerspelling of words from the spoken language. According to
these authors, borrowings from English can enter American Sign Language (ASL) in
various ways: one strategy consists of replacing the native sign with the fingerspelled
shape of the first letter of the English word. Thus room and office are distinguished in
ASL in that the hand takes the shape of the letters R and O respectively but the movement and location of the sign remain the same. An alternative strategy is to fingerspell
loans. Thus the ASL word no has evolved from finger-spelled N O. Insights from sign
language are often overlooked in studies of language contact, and further research into
this area is very much needed.

5. The integration of borrowings

The phonological integration of loanwords has received much more attention in the
literature than the morphological, syntactic or semantic integration. The following
overview gives an idea of the issues that have been discussed.
In his typology of borrowing, which is based on the notions of substitution and
importation, Haugen (1950:214) postulates a morphemic substitution which operates
independently of the phonemic. Other researchers have subsequently corroborated

23

24

Jeanine Treffers-Daller

this assumption. Appel and Muysken (1987:153) state that if one assumes that the
lexicon and the phonological component of the grammar are independent, the meaning
and the phonetic form of a word are not necessarily always borrowed together. Sometimes the entire phonetic form of a word is borrowed, and sometimes it is partly or
entirely substituted by borrowing language sounds. Van Coetsem (1988:8) reserves
the term imitation for the former and adaptation for the latter. Thus, when speakers of
English pronounce the French expression dj vu, they may or may not be successful
in realising the French front rounded vowel [], which does not belong to the inventory of English phonemes. Many speakers will substitute [] with native [u], which is
an example of adaptation in Van Coetsems terminology.
Haugen (1950: 222), Poplack, Sankoff and Miller (1988) and Thomason and
Kaufman (1988) show that there is a lot of variability in the pronunciation of loanwords and that this variation correlates with speaker characteristics such as age and
bilingual ability. Older speakers who have a less elaborate command of the source
language phonology integrate the loanwords to a larger extent into the phonological
patterns of the borrowing language than younger speakers do. Poplack and Sankoff
(1984) and Poplack, Miller and Sankoff (1988) provide evidence for the fact that
phonological integration proceeds as a function of the social integration of the loanword. Widespread loans which have entered the borrowing language at an early date
are often realised with borrowing language pronunciation, whereas the pronunciation
of more recent and less widespread ones is often more similar to the pronunciation in
the source language.
Thomason and Kaufman (1988:124) demonstrate that the importation of large
numbers of loanwords into a language does not necessarily have important consequences for the phonological system of the borrowing language. Although lexical
influence of French on English was very heavy, there is very little structural interference
from French. French loanwords did not introduce any new phones at all into English,
according to Thomason and Kaufman, even though formerly allophonic distinctions,
such as the distinction between [f] and [v], were phonemicized in Middle English
under the influence of French. In other language contact situations the phonological
system of the borrowing language can be changed dramatically, as the case of Asia
Minor Greek (Dawkins 1916, in Thomason & Kaufman 1988) illustrates.
As far as the syntactic integration of loanwords is concerned, gender allocation to
borrowed nouns is a well explored area (Baetens Beardsmore 1971; Chirsheva 2009;
Poplack et al. 1982; Wawrzyniak 1985; Poplack, Sankoff & Miller 1988). Morphological
integration has been studied by Miller (1997), who focuses on the combination of
French derivational suffixes and English roots and vice versa. Others have investigated
the addition of inflectional morphology, such as the formation of plurals of borrowed
words (Poplack et al. 1988; Treffers-Daller 1999).

Borrowing

6. Constraints
The main thrust of research at the end of the seventies and the eighties was to discover
the constraints on the borrowing process. It was clear that most bilingual data contained loanwords of different categories and that some categories were more likely to
be borrowed than others. As mentioned above, all researchers recognise the existence
of lexical borrowing and this is certainly the wide-spread form of borrowing in the
languages of the world. As Weinreich (1953:56) puts it, the vocabulary of a language,
considerably more loosely structured than its phonemics or its grammar, is beyond
question the domain of borrowing par excellence. In addition, Weinreich points to
the socio-cultural reasons behind the fact that the lexicon is so receptive to borrowing: speakers often want to introduce new concepts or try to avoid homonyms and
frequently replace outworn expressions with new ones.
The aim of the studies into constraints on borrowing was to explain why certain categories were more likely to be borrowed than others. Appel and Muysken
(1987:170171) give an overview of different approaches to this problem, and discuss
the principles behind the so-called hierarchies of borrowability (also called scale of
adoptability, cf. Haugen 1950).
One of the earliest hierarchies dates back to the Sanskritist William Dwight
Whitney, who formulated the following hierarchy Whitney (1881), which was taken
up and subsequently elaborated by Haugen (1951).

(1) Nouns other parts of speech suffixes inflection sounds

Muysken (1981a) formulates the following hierarchy on the basis of his analysis of
Spanish borrowings in Quechua:

(2) nouns adjectives verbs prepositions co-ordinating conjunctions quantifiers

determiners free pronouns clitic pronouns subordinating conjunctions

Appel and Muysken (1987:172) explain that paradigmatic and syntagmatic coherence
relations in language are at the basis of these hierarchies. As the pronoun system of a
language is more tightly organised than the adjectives, for example, pronouns are less
likely to be borrowed. Categories that are firmly embedded in the syntagmatic relations in a sentence, such as verbs, are less likely to be borrowed than elements such as
nouns, which are less crucial to the organisation of the sentence.
Moravcsik (1978) formulates the constraints on borrowing in the form of implicational universals of borrowing. Matras (1998:283) summarises and reinterprets these
universals as follows: elements that show structural autonomy and referential stability are more likely to be affected by contact than those that display stronger structural dependency and referential vagueness or abstractness.

25

26

Jeanine Treffers-Daller

The borrowing scale formulated by Thomason and Kaufman (1988) further

develops the ideas advanced in earlier studies by including structural features in the
hierarchy. Where other researchers hesitated whether rules of the grammar can be
imported into another language, Thomason and Kaufman (1988: 14) state that as
far as the strictly linguistic possibilities go, any linguistic feature can be transferred
from any language to any other language. Aikhenvald (2006: 2) confirms this view
and claims that no feature, be it a form or a pattern is entirely borrowing-proof.
In Thomason and Kaufmans model, the crucial social factor determining the extent
and depth of borrowing is intensity of contact: greater intensity of contact or greater
cultural pressure on borrowing-language speakers results in more borrowing. In addition
to intensity of contact, typological distance between languages seems to be relevant for
predicting how much, and what kinds of interference will occur in a borrowing situation (Thomason & Kaufman 1988:72). The principle behind the borrowing scale is
explained as follows:
Our tentative borrowing probability scale, then, is a hierarchy determined by the
relative degrees of structuredness of various grammatical subsystems: the more
internal structure a system has, the more intense the contact must be in order to result
in structural borrowing. We propose that, in the absence of a close typological fit
between particular source-language and borrowing-language structures, features lower
on the scale will not be borrowed before features higher on the scale are borrowed.
(Thomason & Kaufman 1988:734).

At the lowest level in the scale, borrowing is limited to the lexical level and mainly to
content words. Structural borrowing is found at the higher levels only. According to the
scale, the existence of structural borrowing in a language generally implies that words
have also been borrowed. The type and amount of structural borrowing increases
with the amount of contact/pressure exerted on borrowing-language speakers. At
level two, one also finds some borrowed function words, such as conjunctions and
various adverbial particles. In Brussels Dutch, for example, where borrowing is mainly
limited to level 2 phenomena (Treffers-Daller 1999), one finds French adverbs and
adverbial phrases such as donc so, dailleurs anyway, otherwise, furthermore, surtout
especially, peu prs about and de temps en temps from time to time as well as
French conjunctions such as tandis que while and moins dat (French moins que,
unless. An example of slight structural borrowing in this language contact situation
can be found in the adoption of French phonemes, but their occurrence is restricted
to loanwords. Thus, the French nasal vowels are imported in some loan words, such
as donc and de temps en temps. Syntactic borrowing is very limited in Brussels Dutch.
Borrowed adverbial particles appear in a pre-clausal position, rather than in the
first position in the main clause and this phenomenon can perhaps be attributed to
influence from French (Treffers-Daller 1994, 1999).

Borrowing

At level three, one finds borrowing of prepositions and slight structural borrowing.
Derivational affixes may be abstracted from borrowed words and added to native
vocabulary. Miller (1997) in his study of borrowed suffixes on native bases in Middle
English, comes to the conclusion that French suffixes were productive in Middle English,
as there were at least 100 hybrid forms consisting of an English base and a French suffix in
English prior to 1450. The most productive of these were: -able as in understandable, -ess
as in murderess, and -ery as in husbandry.
At levels four and five, major structural features are borrowed. According to
Thomason and Kaufman, Weinreichs (1953) examples of German borrowing in
Romansh, which has lost gender in predicative adjectives and in which the nounadjective word order is partially replaced by adjective-noun order, is probably an example of borrowing at level four. Thomason and Kaufman extensively discuss Dawkins
(1916) analysis of Turkish influence on Asia Minor Greek, which they see as an example
of structural borrowing at level five. In these varieties of Greek, various word order
features are borrowed, as well as vowel harmony. In addition, several grammatical
categories that Turkish lacks, such as gender and adjective-noun agreement, are lost.
Bakker and Mous (1994) and Thomason (1996) convincingly show that there is a
difference between languages which are characterised by heavy borrowing and genuinely mixed languages. Bakker and Mous define mixed languages as a combination
of the grammatical system of one language with the lexicon of another. An example is
Media Lengua, which is a combination of Quechua grammar with a Spanish lexicon
(Muysken 1981b, 1994, 1996). Michif is a different case, because it has the nominal
grammar from French and the verbal grammar from Cree (Bakker 1996). English does
not classify as a mixed language, despite the fact that 75% of its vocabulary is from
French, because the basic vocabulary of English remains almost completely English.
According to Thomason and Kaufman (1988) only 7% of the basic vocabulary of English
is borrowed. In mixed languages such as Media Lengua, almost all words have been
replaced, including the basic vocabulary.
As borrowing and interference through shift can occur in the same contact situation, it is sometimes difficult to establish with certainty whether the phenomena under
consideration are due to one or the other process. Especially in those situations characterised by heavy borrowing or heavy substratum interference, the distinction between
both processes may become blurred. Winford (2003:83) considers it problematic to
treat contact between Asia Minor Greek and Turkish as a case of borrowing in the
strict sense, because the speakers of Asia Minor Greek may not have been the sole
agents of change in this case.
In recent years, several studies have highlighted the structural effects of borrowing.
Aikhenvald and Dixon (2006) and Matras and Sakel (2007) provide an overview of
structural borrowing in a wide range of languages. The latter also introduce the distinction between matter replication (borrowing of morphological material and its

27

28

Jeanine Treffers-Daller

phonological shape) and pattern replication: the organisation, distribution and mapping of grammatical or semantic meaning, while the form itself is not borrowed (Sakel
2007:15). Salmons (1990) is one of the first studies on borrowing in which the focus
shifts from grammar to the level of pragmatics and discourse. Salmons shows that
speakers of American German dialects use mainly English discourse markers (you
know, well, of course) even though literal translations of some German particles and
similar items are being used in expressions such as lets go to town once, come here
once or that too yet (German das auch noch) (Salmons 1990:473). In addition, speakers maintain discourse routines for greetings, leavetakings, etc., which appears to be
common even for speakers who are barely bilingual. Other authors who have focussed
on pragmatic and discourse-analytical analyses of switched and borrowed discourse
markers are Auer (1984, 1998), Fuller (2001), Hlavac (2006), Meeuwis and Blommaert
(1998) and Moyer (1998). According to Stolz and Stolz (1996; in Matras 1998:289),
borrowing begins at the level of discourse and gradually makes its way into clause
combining devices until the level of word grammar.
Matras (1998, 2000) approach to the constraints problem is interesting because he
integrates pragmatic and psycholinguistic considerations in the analysis of constraints
on borrowing. Matras shows that structure-oriented borrowability hierarchies on their
own cannot give a satisfactory explanation of why utterance modifiers (interjections,
discourse markers such as well and anyway, and adverbial particles) are borrowed so
frequently in many contact situations (Salmons 1990; De Rooij 1996; Maeschler 1998).
Although syntactic considerations, such as the lack of syntactic integration of these
elements clearly play a role here, in a comparative study of borrowing in different
dialects of Romani, Matras shows that a number of pragmatic principles explain
the different degrees of borrowability of utterance modifiers which have the same
structural-syntactic status. Instead of aiming for a structural syntactic approach to
a hierarchy of borrowability, Matras develops a function-related hierarchy in which,
for example, contrastive conjunctions are very frequently borrowed, but temporal
conjunctions much less frequently. Matras argues that the process responsible for
the replacement of the utterance modifiers is fusion, i.e. the nonsepararation of the
two systems, and he points to psycholinguistic factors, such as the cognitive pressure
exerted on bilinguals to draw on the resources of the pragmatically dominant language
for situative, gesturelike discourse-regulating purposes (Matras 1998:321) to explain
the hierarchy. In a follow-up study Matras (2000) develops the cognitive model for
bilingual discourse markers further and shows how bilingual speakers accidentally
produce discourse markers from language B while aiming to speak language A, which
is interpreted as evidence for the existence of fusion.
Although pragmatic and discourse-functional perspectives on borrowing remain
rare, Prince (1988) is one of the very few studies which address the issue of pragmatic borrowing, i.e. the borrowing of a discourse function of a particular syntactic
form from another language. Prince shows that the discourse function of Yiddish

Borrowing

dos-sentences, exemplified in (3), was borrowed from Russian, on the analogy of the
Russian expletive eto this:
(3) Dos shlogst du di puter?
This beats you the butter
Its you whos churning the butter?

(Prince 1988:506)

Dos-sentences are functional equivalents of English cleft-sentences, but syntactically

they are the same as es-sentences, such as (4):
(4) Es dremlen feygl oyf di tsvaygn
it doze
birds on the branches
Birds are dozing on the branches.

(Prince 1988:510)

As the syntactic structures of (3) and (4) are the same, Prince argues that (3) is not an
example of syntactic borrowing, but an example of pragmatic borrowing of a discourse
function. It is highly interesting to see that it is possible to borrow a pragmatic function, but unfortunately there are hardly any other examples of pragmatic borrowing
in the literature.
Studies of discourse which include a focus on borrowing remain rare. In his analysis of job applications by semi-literate speakers of French in Cameroon Mboudjeke
(in press) shows how speakers make use of a borrowed discourse strategy: speakers
unconsciously borrow a persuasion strategy from their L1 when speaking French, as
in (5), where the speakers express admiration for the addressee before making their
request, which is common in traditional cultures of Cameroon.

(5) Jai linsigne honneur de venir trs respectueusement auprs de votre haute
et humiliante personnalit solicit un poste de travail dans votre socit en
qualit de chaufeur.
I have the distinguished honour to come before your high and humiliating
personality to apply for the position of a driver in your company.

7. Quantitative approaches
Parallel to the development of the qualitative approaches to borrowing, towards the
end of the eighties, the first quantitative studies of borrowing appeared. This was certainly to a large extent due to the fact that corpora became available in electronic form
and the techniques for the analysis of bilingual data improved tremendously. As a result,
it became possible for researchers to study far larger corpora and to investigate the
quantitative properties of borrowing in more detail than had been possible previously.
Poplack, Miller and Sankoff (1988) were the first to present detailed quantitative analyses of borrowing and to make a comparison between the distribution of loanwords
(established borrowings and nonce borrowings) and native words in a corpus. This

29

30

Jeanine Treffers-Daller

comparative analysis made it possible for them to make well-founded claims about
the likelihood for different words to be borrowed. Their French-English corpus, which
consists of two and a half million words collected in Ottawa-Hull, is still probably
the largest database of bilingual speech collected to date. The results by and large
confirmed the predictions of the hierarchies of borrowability formulated earlier, and
clearly demonstrated that nouns are by far the most frequently borrowed elements,
followed by verbs, adjectives and conjunctions. Interjections and frozen expressions,
which were not taken up in any of the hierarchies of borrowability formulated so far,
formed the third largest category of borrowed words. Prepositions were borrowed less
frequently than predicted by many of the hierarchies and non-lexical categories were
exceedingly rare in the corpus. In total, the English-origin borrowings in French as
spoken in Ottawa-Hull formed less than 1% of the total words in the corpus, showing
that borrowing is a relatively rare phenomenon in this data base. Treffers-Daller (1994)
found similar low rates of borrowing in a French-Dutch corpus of 190.000 words and
showed that there was a clear asymmetry in the directionality of borrowing: French
words were far more likely to be borrowed into Dutch than Dutch words into French.
Using advanced statistical techniques, van Hout and Muysken (1994) study the
influence of four sets of factors on borrowability in a Spanish-Quechua bilingual corpus: lexical content, frequency, structural coherence factors and equivalence. They
show that structural coherence factors, in particular paradigmaticity (whether or
not an item is part of a structural paradigm), are the strongest structural factors in
their model. In other language pairs, other factors may however turn out to be more
important, according to the authors.
Apart from addressing the constraints problem, Poplack et al. also investigated
the social embedding of borrowing in Ottawa-Hull, by making a detailed analysis of
the influence of social factors such as age, gender, social background and residential
area on the frequency and the type of lexical borrowing in Ottawa-Hull. One of the
most important results of this analysis is that environmental factors (the norms of the
community) and not an individuals language proficiency in both languages are the key
predictors of borrowing patterns. Their conclusion is therefore that behavior with
respect to use of borrowings is acquired and not merely a function of lexical need.
(Poplack, Sankoff & Miller 1988:98).

8. Psycholinguistic approaches
Towards the end of the eighties, borrowing and code-switching started to be studied
from a psycholinguistic point of view. From this perspective, the focus is on language
processing (production and perception) of bilingual speech in bilinguals. Grosjean
(1997:232) notes that:

Borrowing

although code-switching and borrowing have continued to be studied extensively

by linguists and sociolinguists the past few years () there is surprisingly little
experimental work in the domain of bilingual language production.

From the pioneering work done by Grosjean and associates (Soares & Grosjean 1984;
Grosjean & Soares 1986; Grosjean 1988, 1995, 1997), it has become clear that important
new insights into the characteristics of borrowing and the differences between borrowing and other language contact phenomena can be gained from psycholinguistic
approaches. Grosjean (1985 et seq) introduced the concept of language modes, which
has subsequently been shown to be a very powerful explanatory concept:
Bilinguals find themselves in their everyday lives at various points along a situational
continuum that induce different language modes. At one end of the continuum,
bilinguals are in a totally monolingual language mode, in that they are interacting
with monolinguals of one or the other of the languages they know. At the other
end of the continuum, bilinguals find themselves in a bilingual mode, in that they
are communicating with bilinguals who share their two (or more) languages and
with whom they normally mix languages (i.e. code-switch and borrow) (Grosjean
1997:228).

When researchers control for the language mode their subjects are in, they have a
better chance of disentangling the different language contact phenomena found in the
speech of bilinguals, and this may in turn help to identify the (differences and similarities
between) the constraints on the various phenomena.
Whereas in a bilingual language mode, all bilingual language phenomena can occur
(interferences, code-switches, borrowings of various types, etc.) this is not the case in a
monolingual language mode. Here code-switches and borrowings are either inexistent
or are usually kept to a strict minimum so as to ensure adequate communication
(Grosjean 1998:228).

Grosjean (1997) also discusses evidence for the fact that borrowing and code-switching
are processed differently.
The concept of language modes is one of the basic characteristics of Grosjeans
interactive activation model of word recognition in bilinguals, named BIMOLA (Bilingual Mode of Lexical Access), developed over many years of experimental research
(see Grosjean 1997 for a detailed presentation).
Greens (1998) Inhibitory Control Model is a model of bilingual speech processing
which aims at explaining how bilinguals control their two languages, for example when
translating from L1 to L2 without actually using L1 words. This is done by assuming that
lemmas are specified in terms of a language tag. Thus, each lemma has an associated
language tag and this tag is one of the factors which affects the activation of the lemma.
After lemmas have been linked to lexical concepts, the model allows for lemmas with
the wrong tag to be inhibited, so that they cannot catch speech production during a

31

32

Jeanine Treffers-Daller

translation task. Though the model was not developed to account for borrowing and
code-switching, it may well have interesting implications for the analysis of language
contact phenomena. One of the issues to be investigated further is whether the concept
of language tag(s) attached to the lemmas of words can be helpful in distinguishing
established borrowings, nonce borrowings and code-switches. It is well known that
many established borrowings can no longer be recognized as borrowings. Thus, native
speakers of English do not recognize people as a borrowing from French, even if they
are fluent speakers of French. Guest words that contain strong language phonetic or
phonotactic cues, such as snob in French, are probably still recognizable. One may
wonder whether in the process of borrowing language tags are lost or replaced and
whether language tags play a role below or beyond the word level (Treffers-Daller 1998).
As detailed phonetic analyses of individual segments and suprasegmental features are
now possible with the help of software such as PRAAT (http://www.fon.hum.uva.nl/
praat/), new insights into the phonetic characteristics of borrowings and other language contact phenomena may well become available in the near future, and these may
help to establish whether it is necessary and feasible to distinguish different contact
phenomena from each other either receptively or productively.

References
Aikhenvald, A.Y. (2006). Grammars in Contact: A cross-linguistic perspective. In A.Y. Aikhenvald &
R.M.W. Dixon (eds.). Grammars in Contact: A Cross-Linguistic Typology: 166. Oxford: Oxford
University Press.
Aikhenvald, A.Y. & R.M.W. Dixon (2006). Grammars in Contact: A Cross-Linguistic Typology. Oxford:
Oxford University Press.
Appel, R. & P. Muysken (1987). Language contact and bilingualism. Arnold.
Auer, P. (1984). Bilingual conversation. John Benjamins.
(ed.) (1998). Code-switching in conversation. Language, interaction and identity. Routledge.
Backus, A. & M. Dorleijn (2009). Loan translations versus code-switching. In B.E. Bullock &
A.J.Toribio (eds.)The Cambridge Handbook of Linguistic Code-switching (pp. 7594). Cambridge:
Cambridge University Press.
Baetens Beardsmore, H. (1971). A gender problem in a language contact situation. Lingua 27: 141159.
Bakker, P. (1996). A language of our own. The genesis of Michif the mixed Cree-French language of
the Canadian Mtis. Oxford University Press.
Bakker, P. & Mous (eds.) (1994). Mixed languages. 15 Case studies in language intertwining. IFOTT.
Blom, J.-P. & J.J. Gumperz (1972). Social meaning in linguistic structure: code-switching in
Norway. In J.J. Gumperz & D. Hymes (eds.) Directions in sociolinguistics: 407434. Holt,
Reinhart & Winston.
Bloomfield, L. (1933). Language. Holt, Reinhart & Winston.
Chirsheva, G. (2009). Gender in Russian-English code-switching. International Journal of Bilingualism
13 (1), 2093.

Borrowing

Clyne, M. (1967). Transference and triggering. The Hague.

Clyne, M. (2003). Dynamics of language contact. Cambridge: CUP.
Colantoni, L. & J. Gurlekian (2004). Convergence and intonation: Historical evidence from Buenos
Aires Spanish. Bilingualism: Language and Cognition 7(2): 107119.
Dawkins, R.M. (1916). Modern Greek in Asia Minor: a study of the dialects of Slli, Cappadocia and
Phrasa with grammars, texts, translations and glossary. Cambridge University Press.
De Rooij, V.A. (1996). Cohesion through contrast: Discourse structure in Shaba Swahili/French
conversations. IFOTT.
Extra, G. & L. Verhoeven (eds.) (1993). Immigrant languages in Europe. Multilingual Matters.
Fuller, J.M. (2001). The principle of pragmatic detachability in borrowing: English-origin discourse
markers in Pennsylvania German. Linguistics 39(2): 351369.
Green, D. (1998). Mental control of the bilingual lexico-semantic system. Bilingualism: Language and
Cognition 1(2): 6781.
Groot, A. De & J.F. Kroll (eds.) (1997). Tutorials in Bilingualism. Psycholinguistic perspectives.
Lawrence Erlbaum.
Grosjean, F. (1985). The bilingual as a competent but specific speaker-hearer. Journal of Multicultural
and Multilingual Development 6: 467477.
(1988). Exploring the recognition of guest words in bilingual speech. Language and Cognitive
Processes 3: 233274.
(1995). A psycholinguistic approach to code-switching: the recognition of guest words by
bilinguals. In L. Milroy & P. Muysken (eds.): 259275.
(1997). Processing mixed language: issues, findings, and models. In A. de Groot & J.F. Kroll
(eds.) Tutorials in Bilingualism. Psycholinguistic perspectives: 225254. Lawrence Erlbaum.
Grosjean, F. & C. Soares (1986). Processing mixed language: some preliminary findings. In J. Vaid (ed.)
Language processing in bilinguals: psycholinguistic and neuro-psychological perspectives: 145179.
Lawrence Erlbaum.
Haspelmath, M. & U. Tadmor (2009). Loanwords in the worlds languages: a comparative handbook.
Berlin: Mouton de Gruyter.
Haugen, E. (1950). The analysis of linguistic borrowing. Language 26: 210231.
Johanson, L. (1993). Code-copying in immigrant Turkish. In G. Extra & L. Verhoeven (eds.) Immigrant
languages in Europe: 197221. Multilingual Matters.
Maeschler, Y. (1998). On the transition from code-switching to a mixed code. In P. Auer (ed.): 125149.
Matras, Y. (1998). Utterance modifiers and universals of grammatical borrowing. Linguistics
36(2): 281331.
(2000). Fusion and the cognitive basis for bilingual discourse markers. International Journal
of Bilingualism 4(4): 505528.
Matras, Y. and J. Sakel (2007). Grammatical Borrowing in Cross-Linguistic Perspective. Berlin/
NewYork: Mouton de Gruyter.
Mboudjeke, J.-G. (in press, 2010). Linguistic politeness in job interviews in Cameroon. Journal
of Pragmatics.
Meeuwis, M. & J. Blommaert (1998). A monolectal view of code-switching: Layered code-switching
among Zairians in Belgium. In P.Auer (ed.): 76100.
Miller, G.D. (1997). The morphological legacy of French. Borrowed suffixes on native bases in Middle
English. Diachronica 14(2): 233264.
Milroy, J. (1997). Internal versus external motivations for linguistic change. Multilingua 16(4): 311323.
Milroy, L. & P. Muysken (eds.) (1995). One speaker two languages. Cambridge University Press.

33

34

Jeanine Treffers-Daller
Moravcsik, E.A. (1978). Language Contact. In J. Greenberg (ed.) Universals of Human Language:
93122. Stanford University Press.
Moyer, M. (1998). Bilingual conversation strategies in Gibraltar. In P. Auer (ed.) 215236.
Muysken, P. (1981a). Quechua en Spaans in het Andesgebied. [Quechua and Spanish in the Andes.]
Tijdschrift voor Taal- en Tekstwetenschap 1: 124138.
(1981b). Halfway between Quechua and Spanish: the case for relexification. In A. Highfield &
A.Valdman (eds.) Historicity and Variation in Creole Studies, 5278. Karoma.
(1990). Concepts, methodology and data in language contact research: Ten remarks from the
perspective of grammatical theory. In Papers for the workshop on concepts, methodology and
data, Network on Code-switching and language contact: 1530. Basel.
(1994). Media Lengua. In P. Bakker & M. Mous (ed.): 207211.
(1996). Media Lengua. In S.G. Thomason (ed.): 365426.
Pfaff, C. (1979). Constraints on language mixing. Language 55: 291318.
Poplack, S. (1980). Sometimes Ill start a sentence in Spanish Y TERMINO EN ESPANOL: toward a
typology of code-switching. Linguistics 18: 581618.
Poplack, S. & D. Sankoff (1984). Borrowing: the synchrony of integration. Linguistics 22: 99135.
Poplack, S., A. Pousada & D. Sankoff (1982). Competing influences on gender assignment: variable
process, stable outcome. Lingua 57: 128.
Poplack, S., D. Sankoff & C. Miller (1988). The social correlates and linguistic processes of lexical
borrowing and assimilation. Linguistics 26: 47104.
Poplack, S. & M. Meechan (1995). Patterns of language mixture: nominal structure in Wolof-French
and Fongbe-French bilingual discourse. In L. Milroy & P. Muysken (eds.): 199232.
Prince, E.F. (1988). On pragmatic change: the borrowing of discourse functions. Journal of Pragmatics
12: 505518.
Queen, R.M. (2001). Bilingual intonation patterns: Evidence of language change from Turkish-German
bilingual children. Language in Society 30(1): 5580.
Romaine, S. (1995). Bilingualism. 2nd Ed. Blackwell.
Sakel, J. (2007). Types of loan: Matter and pattern. In Y.Matras & J. Sakel. (eds.) Grammatical borrowing
in cross-linguistic perspective (pp. 1529). Berlin/New York: Mouton de Gruyter.
Salmons, J. (1990). Bilingual discourse marking: code switching, borrowing and convergence in
some German-American dialects. Linguistics 28: 475492.
Salverda De Grave, J.J. (1906). De Fransche woorden in het Nederlands. [French words in Dutch.]
Verhandelingen der Koninklijke Akademie van Wetenschappen te Amsterdam. Afdeeling
Letterkunde, Nieuwe Reeks, deel 6.
Sandler, W. & D. Lillo-Martin (2006). Sign Language and Linguistic Universals. Cambridge, UK:
Cambridge University Press.
Soares, C. & F. Grosjean (1984). Bilinguals in a monolingual and a bilingual speech mode: the effect
of lexical access. Memory and Cognition 12: 380386.
Stolz, C. & T. Stolz (1996). Funktionswortentlehnung in Mesoamerika. Spanish-Amerindischer
Sprachkontakt (Hispanoindia II). Sprachtypologie und Universalienforschung 49(1): 86123.
Thomason, S.G. (ed.) (1996). Contact languages. A wider perspective. John Benjamins.
Thomason, S.G. & T. Kaufman (1988). Language contact, creolization and genetic linguistics. University
of California Press.
Treffers-Daller, J. (1994). Mixing two languages: French-Dutch contact in a comparative perspective.
Mouton de Gruyter.
(1998). The IC model and code-switching. Bilingualism: Language and Cognition 1(2): 9899.

Borrowing

(1999). Borrowing and shift-induced interference: contrasting patterns in French-Germanic

contact in Brussels and Strasbourg, Bilingualism: Language and Cognition 2: 122.
(2005). Code-switching. In U. Ammon, N. Dittmar, K. Mattheier & P. Trudgill (eds.) Sociolinguistics, An International Handbook of the Science of Language and Society, 2nd ed., vol. 2:
14691482. de Gruyter.
(2009). Code-switching and transfer: an exploration of similarities and differences.
In: B.E.Bullock & A.J. Toribio (eds.)The Cambridge Handbook of Linguistic Code-switching
(pp.5874). Cambridge: Cambridge University Press.
Vaid, J. (1986). (ed.) Language processing in bilinguals: psycholinguistic and neuro-psychological
perspectives. Lawrence Erlbaum.
Van Coetsem, F. (1988). Loan phonology and the two transfer types in language contact. Foris.
Van Hout, R. & P. Muysken (1994). Modeling lexical borrowability. Language Variation and Change
6: 3962.
Wawrzyniak, U. (1985). Das Genus franzsischer Lehnwrter im Deutschen, Zeitschrift fr Sprachwissenschaft 4(2): 201217.
Weinreich, U. (1953). Languages in contact. Mouton.
Weinreich, U., W. Labov & M.I. Herzog (1968). Empirical foundations for a theory of language change.
In W.P. Lehmann & Y. Malkiel (eds.) Directions for historical linguistics: 95188. University of
Texas Press.
Whitney, W.D. (1881). On mixture in language. Transactions of the American Philosophical Association
12: 126.
Winford, D. (2003). An introduction to contact linguistics. Blackwell.

35

Contact linguistics
Michael Meeuwis & Jan-Ola stman

1. Introduction
The purpose of this article is to give an overview of issues related to the processes
behind, and linguistic impact of, various types of contact between languages and
language users, to what happens when languages/language users come into contact,
and when they are (not) in contact. While paying specific attention to issues directly
relevant to pragmatics, the exposition here will be a fairly general survey of existing research on language-contact phenomena, with an attempt to present the often
implicit large-scale framework of study in this field.
Various terms have been used for language-contact phenomena and processes;
we will introduce several of them as we go along. Note, though, that the term contact
linguistics has also sometimes been used in a slightly narrower sense (see Section 3)
than it is used in the title of this article, where it is to designate all linguistic approaches
to contact.
The presentation of the different (socio-)linguistic approaches to language contact, and of the variety of phenomena that are documented to be involved in cases of
language contact will be organized under three headings: the location, the direction,
and the process of contact. Before that, we wish to locate the study of language contact
more broadly with respect to approaches concerned with the typology and classification
of languages.

2. Contact in relation to classificatory bases in linguistics

Languages have been classified in various ways over the centuries, but too seldom
attempts have been made to approach the different kinds of classifications systematically
with a view to integrate them.
One set of classifications takes the particularities of individual languages as a
starting point, and focuses on how each language has come to look the way it does.
Two main types of explanation are commonly advanced: (i) genetic relationships
among languages and (ii) areal relationships among languages. The problem of what
is the result of genetic relationship and what similarities are due to contact situations
directly or indirectly received extensive attention all through the 20th century.

Contact linguistics

Another set of classifications focuses more on language as a theoretical concept,

on what categories and distinctions are actually made use of in natural languages, out
of all theoretical possibilities. This kind of research focuses on (i) language universals
and (ii) language typologies. The question of what languages have in common has
always interested humans, since such a faculty of language for long seemed to separate
humans from animals. In this respect, theoretical and systematic studies of language
universals and of typological similarities between languages complement each other
in a fruitful manner.
A definitional characteristic of what we consider as human languages is that they
are an integrated part of social contacts. Both phylogenetically and ontogenetically, a
language emerges from contact; equally important, languages come into contact with
each other via speakers and communities, and within speakers. The result of linguistic
contact, viz. linguistic interference or, more generally, linguistic influence, therefore
deserves and requires every linguists attention, regardless of her or his subdiscipline
of specialization.1
3. Contact and location
Because it is both theoretically and practically impossible to separate out which characteristics of a language are due to areal and which to genetic causes, the question as a
whole is historically irrelevant if asked in vacuo, i.e. as related primarily to linguistic
form. The physical location of the settlement of the speakers of a language becomes
interesting only when it is coupled with the culture and ideology of the speakers and
their relation to neighboring communities.
What we regard as linguistically distinctive geographical dialects and languages
have evolved either through long periods of lack of contact between neighboring
communities, or through a process of extensive contact between different language
varieties, mediated by stages of degrees of creolization. In many cases, however, different
dialects or languages live side by side for centuries in a give-and-take relationship, in
some way accommodating to each other. Locations where (a clustering of) linguistic
features have been established as being due to results of such contact processes are

. Although fully aware of the negative connotations attached to the term interference, we prefer
to maintain its use for the present article, as it allows us to distinguish between the adoption in one
language of elements from another language (interference), and the broader sense of influence,
which covers all types of linguistic processes and results by which languages in contact can be
linguistically affected (see Section 5).

37

38

Michael Meeuwis & Jan-Ola stman

often referred to as Sprachbund/Sprachbnde, or linguistic areas, and the branch of

linguistics which studies such phenomena is referred to as areal linguistics.2
Areal linguistics in this traditional sense has obvious connections with dialectology,
geolinguistics, and what is sometimes referred to as contact linguistics proper. However, whereas dialectology is concerned with linguistic variation at large, and areal
linguistics focuses on linguistic convergences in large-scale areas bound together by
a joint culture and history, and geolinguistics in turn is primarily concerned with
what happens to languages that are grouped together, e.g. in a state, contact linguistics
stricto sensu is more interested in contacts between speakers, and what happens in
the language and cognition of individuals. Here, bilingualism and multilingualism are
especially in focus.
As mentioned, areal linguistics focuses traditionally on (groups of) speakers that
are located physically as neighbors. Language contact, however, also significantly occurs
within non-locationally defined linguistic groupings (for which the term Sprechbund is
sometimes used). These groupings are united by similar kinds of interests and features
among its members (age, sex, class, occupation, religion, interests, hobbies, etc.) and
they are linguistically characterized by a social and/or psychological language variety
in relation to the community at large. Due to modern advances in telecommunication
and information technology and to more extensive traveling, members within such
groupings, although not being physically adjacent, come into close contact with one
another, and their languages and communicative behaviors in general take on features
from a joint pool (Paolillo et al. 2005; Androutsopoulos 2006).
4. Direction of interference
A quite different starting point for approaching language contact phenomena is language interference, i.e. the use and, to varying degrees, the integration of linguistic
features or elements of one language in another. On the basis of the sheer amount
of documentation in the literature, one might well deduce that this is the dominant
province of research and theorizing in the field of contact linguistics today.

. The best known and most studied linguistic areas are the Sprachbnde on the Balkan peninsula
(as suggested early on by Trubetzkoy & Jakobson), the Indian continent and neighboring areas
in Southeast Asia (e.g. Emeneau 1956; Masica 1976), and various groupings of American Indian
languages, for example in the western and northwestern USA (e.g. Haas 1969). Interesting borderline cases of linguistic convergence which some linguists are hesitant to qualify as linguistic areas
include the North Sea littoral (cf. e.g. Ureland ed. 1987; & Poussa ed. 2002), the Circum-Baltic
area, aka Baltic Europe (cf. Lehiste 1988; Raukko & stman 1994; Dahl & Koptjevskaja-Tamm eds.
2001), and the Mediterranean (see Ramat & Roma eds. 2007).

Contact linguistics

Haugen (1950), Weinreich (1953), and Clyne (1972) provided a systematic terminological and typological backbone for the 19th-century and early 20th-century intuitions about language interference. In the aftermath of these publications, a commonly
shared assumption emerged, according to which cross-language interferences can
usefully be divided into two global types. A first group of phenomena consists of interferences that occur from what is for a speaker or a community a foreign (FL) or second,
third, etc. language (L2) to her or his first, habitual, or dominant language (L1).3 The
second type of interferences proceeds in the opposite direction: elements of the L1 are
transferred to the L2/FL.
The most explicit manifestation of the usefulness of this distinction is the very
structure of contact linguistics as a field of study. Linguistic interference of the first
type is commonly dealt with in subfields that are mostly community-oriented, such as
the ones referred to in the foregoing section, whereas disciplines such as L2 acquisition
research typically address linguistic interferences of the second type.
One of the appeals of this distinction is a common recognition that these two
types differ significantly in the outcomes of, and the processes and mechanisms at
work in, situations of language contact (Thomason & Kaufman 1988; Van Coetsem
1988; Winford 2003).
The first type of cross-language interference, by some also called borrowing, is no
doubt best known by cases of lexical interference. But borrowing is not only a lexical
phenomenon. As Thomason & Kaufman (1988) and Thomason (2001) argue, even
if universal tendencies and susceptibility hierarchies cannot be ignored, there are
no structural barriers that would preclude any level of linguistic structuring from
phonology to the most complex syntactic patterns (see also Heine & Kuteva 2005 on
the notion of grammatical replication or calquing) to be affected by cross-language
borrowing. Moreover, studies such as Prince (1988) and Valds & Pino (1981) showed
that pragmatic aspects like topicalization features or the organization of speech acts
can also be transferred from an L2/FL to an L1.
The social contexts in which borrowing occurs are manifold. The traditionally
recognized and well-documented contexts of cultural influence by immediate cohabitation or by contact through various possible types of media have received renewed
interest recently in studies on the effects of globalization and on related changes in late
modernity. These studies deal not only with the types of adaptation that take place in
speech and writing when elements and structures are getting integrated in the L1, or
. As is common in second-language acquisition research and related fields, we will in the
following be using L2 to refer not only to second language, but to all languages (L3, L4 Ln)
acquired after or next to the first. Foreign languages differ from L2s in that they do not directly
fulfill any major socially functional role within the speech community of the individual or group
to whom the term applies.

39

40

Michael Meeuwis & Jan-Ola stman

with the forces (of some level of normative purism) that through replacements and the
creation of new native words struggle to keep the L1 intact, but the impact of speakers attitudes and opinions has also been found to be very decisive in these processes.
These issues are dealt with in detail in Kristiansen & Sandy (eds. forthc.) in relation
to the impact of English in northern Europe, showing for instance the importance
of speakers implicit, subconscious attitudes in understanding contact phenomena in
relation to globalization phenomena.
In addition to the traditional contexts and the recent trends with respect to globali
zation and glocalization, which have also seen an increase in studies on English as a
lingua franca (ELF; cf. Seidlhofer et al. 2006), two further specific contexts where
borrowing plays an important role receive attention in the literature: L1 attrition and
language death. Research on L1 attrition (e.g. Hansen & Reetz-Kurashige 1999) involves
the study of the way mostly younger learners of an L2 gradually lose knowledge of
their L1, and in this process transfer elements from the acquired language to the L1.
Language death goes beyond the individual and concerns similar phenomena from
the angle of the community as a whole (e.g. Dorian ed. 1989; Crystal 2000).
The most typical instance of the second type of contact interference phenomena
from L1 to L2/FL occurs in second language acquisition and foreign language
learning (e.g. Larsen-Freeman 2000; Freeman & Freeman 2001; Ringbom 2006). The
community-oriented counterpart to this individual-based discipline examines language shift (e.g. Fishman ed. 1991; Li Wei 1994). At the intersection of these two
disciplines, an important domain of study which also addresses problems of interference from L1 to L2/FL deals with so-called world varietes. The focus of investigation is on how the world-wide spread of mostly Western languages such as English and
French, leads to the definitive incorporation of L1 interference phenomena in these
languages (Bolton & Kachru 2006).
In much the same way as for borrowing, the linguistic features involved in interference from L1 range from lexical items over phonological features to syntactic structures. Further, L2 acquisition research has been particularly rewarding in drawing
attention to the transfer of pragmatic features from L1 to L2/FL. Bou Franchs (1998)
and Barrons (2003) overviews of pragmatic transfer testify to the great importance
L2 acquisition research attaches to this phenomenon. Another important field of study
which has stressed the importance of such pragmatic transfer is to be found in intercultural communication. Especially the interactionalist tradition within this discipline
(e.g. Gumperz 1982) demonstrated that speakers can master all the purely linguistic
(formal) aspects of the target language, appearing in that way to be perfect bilinguals,
while they at the same time stick to their native pragmatic system; this often turns out
to be the cause for misunderstandings in intercultural encounters.
Despite the global significance of the dichotomy between the two major directions
of cross-language interference, it should be noted that not all linguistic processes and

Contact linguistics

products of language contact can straightforwardly be categorized as either borrowing

from L2/FL or transfer from L1. There is, first, the well-documented case of codeswitching. The distinguishing feature that is generally taken to set code-switching
apart from other types of language contact is the fact that the replacement of original structures by foreign linguistic features is not fully generalized, and that these
foreign features remain to some extent phonologically or morphologically unintegrated (e.g.Auer 1998; Deuchar et al. 2007; Poplack 2004). A second case is that of
mixed languages (Mous 2003; Bakker & Matras eds. 2003). A mixed language draws
its grammar from one language and its vocabulary from another. It is most often the
result of speakers deliberate combinations of two languages in their efforts to mark
a new identity. In time, the original, identity-related intentions become socially less
relevant, but the new language remains in place, and is handed over to new generations.
Third, we have cases where the speakers of the two languages in contact use languages
that are to some extent mutually intelligible, and by virtue of the contact koineization
occurs, where a new variety comes about. Koineization (cf. Kerswill 2002) typically
occurs in combination with migration and/or when speakers come together to live in
the vicinity of newly established workplaces. Fourthly, there is the phenomenon called
false friends, where the adoption goes from one L2/FL to another (Ringbom 1987). A
native speaker of German who knows French as an L2, may send some features of this
French to her or his knowledge of Spanish when acquiring it. Other problems with the
dichotomy are related to the fact that a clear identification of what constitutes the L1
and what constitutes the L2 or FL for any given individual or community is not always
feasible. In the case of child bilingualism, for instance, a child learns two languages
at the same time as equivalent first languages, and easily transfers elements from one
language into the other (Lanza 1997).
In language interference a further distinction in terms of the status of the languages in contact also has to be taken into account. The more powerful language,
i.e.the language of the dominant culture, is referred to as the superstratum language
and fulfills the role of donor for the interference, while the less powerful language is
the substratum language and acts as the recipient language. Although this distinction
is helpful as an analytic tool, it should not be taken as explanatory by itself to suggest,
for instance, that superstratum languages and cultures always tend to win linguistically. For one thing, there are clear cases in history which show that this is not the
case. Secondly, in cases such as pidgins and abrupt creoles the new language cannot
readily be genetically affiliated with either one of the languages involved in the contact
situation (Sebba 1997), and any distinction between donor and recipient language or
between substratum or superstratum is only likely to distort the picture. In Section 3,
we discussed similar problems in phenomena like Sprachbnde. In cases like these,
where languages are of equal importance for the creation of a new language, one often
talks about adstratum situations.

41

42

Michael Meeuwis & Jan-Ola stman

5. Indirect influence in language contact

As we argued in the foregoing section, some form of cross-language interference most
typically occurs in language-contact situations. However, other types of processes
and results of language contact also have to be recognized. These types cover cases in
which a certain linguistic change can be said to be informed by a contact situation but
in which this change is not materialized in a feature that can distinctively be traced to
one of the languages involved in the contact situation.
Thus, there is a host of generalization, avoidance, simplification and other processes at work in second-language acquisition, language shift, language death and other
types of language contact, which show no isomorphic or qualitative correspondence
with elements in the donor language or in a previous form of the recipient, affected
language. These seemingly internal creations of previously non-existent structures
occur simply due to the contact situation, i.e. they would not have occurred in the language unless it had been in contact (e.g. Campbell & Muntzel 1989; Thomason 2001)
and could thus not be explained without reference to a linguistic and cultural contact.
Also, in all these fields, the phenomenon of multiple causation or, as it is also called,
synergy has to be taken into account. Synergy refers to changes that occur internally
in a language but are triggered by the contact with similar features, patterns, or rules
in another language. Here, some qualitative resemblances to features in the donor language or in a pre-contact state of the affected language can be observed: the contact
has given rise to a change in existing features. As Dorian (1993) and Woolard (1989)
argued, synergy is a widely observable phenomenon that is too often slighted by
keeping the distinction between internally and externally motivated language change
too rigidly separate.

6. Contact as process: Towards pragmatics

Late modern society offers considerable challenges to our understanding of language
contact and what direction pragmatic research within contact linguistics needs to
take in the future. Attitudes and social change (Coupland 2009) are getting more and
more attention and processes like dialect levelling and regionalization may have
very different causes, albeit that the results from such processes may look the same
(stman2008).
Contact, and interference and influence through contact, are intrinsically dynamic
notions although they are not always seen from this perspective. Meeuwis (1991) and
Meeuwis & Brisard (1993) drew attention to the inherent a-processual and a-historical
nature of the theoretical and methodological assumptions underlying much of the
research in contact linguistics and related fields concerned with diachronic issues.

Contact linguistics

One possible solution to this is to approach contact phenomena from a pragmatic

perspective, which takes fully into account the continuous adaptation processes involved
in contact situations, and thus avoids treating contact-induced language change as consisting of the mere temporal succession of static products. Such a perspective should
also allow for a type of analysis in which language, culture, and mind are seen as virtually
inseparable, which is another prerequisite for a processual, dynamic approach to language contact. This combines the locational and directional/cognitive approaches and
in addition incorporates in the analysis linguistic manifestations of aspects of culture,
ideology, and implicitness. Thus, also, processual analyses of contact phenomena in
this sense fall under the study of language function, i.e. pragmatics.
Raukko & stman (1994) and stman & Raukko (1995) suggested a processual
approach to areal linguistics where the main tenets are precisely that linguistic and
cultural contacts are closely intertwined and that the foundation for a linguistic area
is a cultural area. The socio-cultural, pragmatic aspects are taken as the basis for analysis; in addition to the results of the process, the process itself is in focus. It is also
suggested that implicit pragmatic aspects are among the first to be taken over in a
convergence situation.
Hence pragmatics as the level of structuring at which linguistic features affected
by language contact are situated, and pragmatics as a preferred approach to the study
of all types of language contact, converge.

References
Auer, P. (ed.) (1998). Code-switching in conversation. Routledge.
Androutsopoulos, J. (2006). Introduction: Sociolinguistics and computer-mediated communication.
Journal of Sociolinguistics 10: 419438.
Bakker, P. & Y. Matras (eds.) (2003). The mixed language debate. Mouton.
Barron, A. (2003). Acquisition in Interlanguage Pragmatics. Benjamins.
Bolton, K. & B.B. Kachru (2006). World Englishes: Critical concepts in linguistics. Routledge.
Bou Franch, P. (1998). On pragmatic transfer. Studies in English language and Linguistics 2: 520.
Campbell, L. & M.C. Muntzel (1989). The structural consequences of language death. In N. Dorian (ed.):
181196.
Clyne, M. (1972). Perspectives on language contact. Hawthorn.
Coupland, N. (2009). Dialects, standards and social change. In M. Maegaard et al. (eds.) Language
attitudes, standardization and language change: 2749. Novus.
Crystal, D. (2000). Language death. Cambridge University Press.
Dahl, . & M. Koptjevskaja-Tamm (eds.) (2001). Circum-Baltic languages, vol. 12. Benjamins.
Deuchar, M., P. Muysken & S. Wang (2007). Structured variation in Codeswitching. International
Journal of Bilingual Education and Bilingualism 10(3): 298340.
Dorian, N. (1993). Internally and externally motivated change in language contact settings. In
C.Jones (ed.) Historical linguistics: 131155. Longman.

43

44

Michael Meeuwis & Jan-Ola stman

Dorian, N. (ed.) (1989). Investigating obsolescence. Cambridge University Press.
Emeneau, M.B. (1956). India as a linguistic area. Language 32: 316.
Fishman, J. (ed.) (1991). Reversing Language Shift. Multilingual Matters.
Freeman, D. & Y. Freeman (2001). Between worlds: Access to second language acquisition. Heinemann.
Gumperz, J. (1982). Discourse strategies. Cambridge University Press.
Haas, M.R. (1969). The prehistoiry of languages. Mouton.
Hansen, L. & A. Reetz-Kurashige (1999). Investigating second language attrition. In L. Hansen (ed.)
Second language attrition: Evidence from Japanese contexts: 320. Oxford University Press.
Haugen, E. (1950). The analysis of linguistic borrowing. Language 26: 210231.
Heine, B. & T. Kuteva (2005). Language contact and grammatical change. Cambridge University Press.
Kerswill, P. (2002). Koineization and accommodation. In J.K. Chambers, P. Trudgill & N. Schilling-Estes
(eds.) The handbook of language variation and change: 669702. Blackwell.
Kristiansen, T. & H. Sandy (eds.) (forthc.). The linguistic consequences of globalization in the Nordic
countries. [Thematic issue of The International Journal of the Sociology of Language.]
Lanza, E. (1997). Language mixing in bilingual children. Oxford University Press.
Larsen-Freeman, D. (2000). Second language acquisition. Annual Review of Applied Linguistics
20:165181.
Lehiste, I. (1988). Lectures on language contact. MIT Press.
Li Wei (1994). Three generations, two languages, one family. Multilingual Matters.
Masica, C.P. (1976). Defining a linguistic area. University of Chicago Press.
Meeuwis, M. (1991). A pragmatic perspective on contact-induced language change. Pragmatics
1:481516.
Meeuwis, M. & F. Brisard (1993). Time and the diagnosis of language change. Antwerp Papers in
Linguistics 72.
Mous, M. (2003). The making of a mixed languages. Benjamins.
stman, J-O. (2008). Det globala i det lokala. In J. Lindstrm et al. Svenskan i Finland 10: 2234.
Nordica.
stman, J-O. & J. Raukko (1995). The pragmareal challenge to genetic language tree models. In
S.Suhonen (ed.) The Fenno-Baltic cultural area: 3169. University of Helsinki, Castrenianum.
Paolillo, J. et al. (2005). Measuring linguistic diversity on the internet. UNESCO. http://www.uis.
unesco.org/template/pdf/cscl/MeasuringLinguisticDiversity_En.pdf [last consulted March 6, 2009].
Poplack, S. (2004). Code-switching. In U. Ammon, N. Dittmar, K.J. Mattheier & P. Trudgill (eds.)
Sociolinguistik: 589596. De Gruyter.
Poussa, P. (ed) (2002). Relativization in the North Sea Littoral. Lincom Europe.
Prince, E. (1988). On pragmatic change. Journal of Pragmatics 12: 505518.
Ramat, P. & E. Roma (eds.) (2007). Europe and the Mediterranean as linguistic areas. Benjamins.
Raukko, J. & J-O. stman (1994). Pragmaattinen nkkulma Itmeren kielialueeseen. University of
Helsinki, Dept. of General Linguistics.
Ringbom, H. (1987). The role of the first language in foreign language learning. Multilingual Matters.
(2006). Cross-linguistic similarity in foreign language learning. Multilingual Matters.
Sebba, M. (1997). Contact languages: Pidgins and creoles. Palgrave MacMillan.
Seidlhofer, B., A. Breiteneder & M-L. Pitzl (2006). English as a lingua franca in Europe. Annual
Review of Applied Linguistics 26: 134.
Thomason, S.G. (2001). Language contact: An introduction. Edinburgh University Press.
Thomason, S.G. & T. Kaufman (1988). Language contact, creolization, and genetic linguistics. University
of California Press.

Contact linguistics

Ureland, S. (ed.) (1987). Sprachkontakt in der Hanse. Niemeyer.

Valds, G. & C. Pino (1981). Muy a tus rdenes: Compliment responses among Mexican-American
bilinguals. Language in Society 10: 5372.
Van Coetsem, F. (1988). Loan phonology and the two transfer types in language contact. Foris.
Weinreich, U. (1953). Languages in contact. Linguistic Circle of New York. [Mouton, 1963].
Winford, D. (2003). An introduction to contact linguistics. Blackwell.
Woolard, K.A. (1989). Language convergence and death as social process. In N. Dorian (ed.):
355367.

45

Creoles and creolization

Salikoko Mufwene

Strictly speaking, creoles are new vernaculars which developed out of contacts between
colonial nonstandard varieties of a European language and several non-European
languages in the Atlantic, Indian, and to some extent the Pacific Oceans during the
seventeenth through the early twentieth centuries. Examples include Saramaccan and
Sranan in Surinam (lexified by English, with the former also heavily influenced by
Portuguese and the latter by Dutch), Papiamentu in the Netherlands Antilles (lexified by Portuguese and influenced by Spanish), Haitian, Mauritian, and Seychellois
(lexified by French), and Gullah in the United States, Hawaiian Creole, Jamaican, and
Guyanese (also lexified by English). Hawaiian Creole, locally identified as Pidgin, is
the only vernacular of the Pacific identified as creole. Vernaculars such as Tok Pisin
(spoken in Papua New Guinea) and Bislama (spoken in Vanuatu) are typically labeled
as expanded pidgins (see below).1
Some creolists have extended the term creole also to some other varieties which
developed during the same period out of contacts with primarily non-European languages, regardless of the role which European colonizers may have played in their
developments. Examples include Sango in the Central African Republic, KikongoKituba and Lingala in Zaire and the Republic of the Congo, and Kinubi in Southern
Sudan and in Uganda.2
Claims that creoles lexified by European languages have more or less the same
structural design (Bickerton 1981, 1984; Markey 1982) are as disputable as the counterclaim that they are more similar to each other in the sociohistorical ecologies of their
developments (Mufwene 1986a). Variation in the latter may be correlated with variation in their structural features and the same equation may be extended to other new
contact-induced language varieties (Mufwene 1995, 2005, 2008).
Creoles have traditionally been distinguished from another group of new language
varieties called pidgins, which also developed out of contacts between Europeans and

. The verb lexify is used between scare quotes simply because the assumption that creoles
kept only the overwhelming proportion of their vocabularies from the European language but
developed their grammars from other sources is highly disputable (Chaudenson 2001; Mufwene
2001, 2008; DeGraff 2001, 2009).
. Holm (1989) presents the most comprehensive inventory with brief sociohistorical backgrounds and sample texts of pidgins (see below) and creoles.

Creoles and creolization

non-Europeans outside Europe, e.g. Nigerian Pidgin, Cameroon Pidgin, Tok Pisin
(Papua New Guinea), and Bislama (Vanuatu), all lexified by English. Since the early
20th century, pidgins have been claimed to differ from creoles in that they either have
no native speakers or do not function as vernaculars. Putatively they are simpler and
restricted to limited-contact functions, typically trade communication. Hawaiian Pidgin
English developed out of the sporadic contacts which primarily Asian contract laborers,
who typically evolved in their own ethnic groups, had with Hawaiians and speakers of
other languages, including Americans. Variation in its structures led Bickerton (1981,
1984, 1999) to claim that pidgins, albeit incipient ones, have no syntax. However, the
expanded ones, which now function as vernaculars and are used in as many functions as
creoles, have complex grammatical systems, as is evident from, e.g. Fral (1989).
There are, nonetheless, significant reasons for not lumping expanded pidgins and
creoles in the same category, usually identified as creole and associated with the fact
that they have acquired a native speaker population. As becomes clearer below, they
evolved in opposite directions, although they are all associated with plantation colonies,
with the exception of Hawaiian Creole, which actually evolved in the city (Roberts 1998,
2005; Mufwene 2004, 2008). Creoles started from closer approximations of their lexifiers and then evolved gradually into basilects that are morphosyntactically simpler,
in more or less the same ways as their lexifiers had evolved from morphosyntactically
more complex Indo-European varieties, such as Latin or Old English. According to
Chaudenson (2001), they extended to the logical conclusion a morphological erosion
that was already under way in the nonstandard dialects of European languages that
the non-Europeans had been exposed to. On the other hand, expanded pidgins started
from rudimentary language varieties that complexified as their functions increased
and diversified. The historical considerations given below shed light on this contrast.
Like creole, the term pidgin has been extended to varieties which developed out of
contacts of primarily non-Europeans, for instance, Chinook Jargon (USA), Delaware
Pidgin (USA), and Hiri Motu (Papua New Guinea).3
The term creole was originally coined in Iberian colonies, apparently in the
16thcentury, for descendants of Iberians and Africans born in the American colonies.
(See Mufwene 1997 and Chaudenson 2001, 2003 for references.) It was adopted in
metropolitan Spanish and later in French toward the end of the century. It was then
borrowed into English in the early 17th century and extended during the second
. Although Mhlhusler (1986) distinguishes between Jargon and Pidgin, the terms are often
confused in the literature (Mufwene 1997, 2005a). Jargon in Chinook Jargon stands for pidgin. In
referring to particular varieties as patois or jargon, the European colonizers just used these terms
as in traditional laymans parlance in Europe for varieties that they did not understand and considered inferior. Unfortunately linguists have adopted them in their technical language uncritically
and in ways inconsistent with colonial history, as shown below.

47

48

Salikoko Mufwene

half of the same century (in all the relevant European languages) to descendants of
Africans and Europeans in other colonies. It was also used as an adjective to characterize
plants, animals, and customs typical of the same territories (Valkhoff 1966; Chaudenson
1992, 2001, 2003).
Creole may not have applied widely to language until the late 18th century, though
Arveiller (1963:207) cites La Courbes Premier voyage (1688: 192) in which it is used
for corrupted Portuguese spoken in Senegal. Such usage may have been initiated
by metropolitan Europeans to disfranchise particular colonial varieties of their languages. Holm (1988: 19) contains an interesting quotation from Oldendorp (1777)
speaking of what is translated as Crioles language. Interestingly, as pointed out by
Chaudenson (1992, 2001, 2003), the earliest Creole people of the homestead phase
spoke non-creole vernaculars that were closer approximations of the relevant lexifiers.
To date, in places such as Martinique and Louisiana, Whites who identify themselves
as Creoles deny that they speak Creole. Likewise, there are various populations in
former Spaniard colonies who are identified as Creoles but speak no particular creoles,
as none really developed in these territories. On the other hand, several speakers of
creoles, especially in former Anglophone colonies, have been told only by linguists
that their vernaculars are called creoles and are considered separate languages from
English (Mhlhusler 1985; Mufwene 1988).
Focusing on language, we note that up to the end of the 19th century, there was
no developmental connection between creoles and pidgins. One of the first connections
is claimed by Bloomfield (1933: 474): when the jargon [i.e. pidgin] has become the
only language [i.e. vernacular] of the subject group, it is a creolized language. Hall
(1962, 1966) reinterpreted this, associating the vernacular function of creoles with
nativization. He then also introduced the pidgin-creole life-cycle to which DeCamp
(1971) added a post-creole stage. To be sure, such claims that creoles have evolved
from pidgins were prompted by the fact that creoles (and expanded pidgins) are
structurally more complex than pidgins and by the common assumption that structures evolve from simpler to more complex ones. Had linguists remembered that the
European lexifiers had actually evolved from earlier, more complex morphosyntaxes,
they would have probably taken into account more of the socioeconomic histories
of the colonies, which suggest a parallel evolution. Indeed, as Baker & Mhlhusler
(1990) point out, the term pidgin was first attested in English (before it was used in
any other European language) in 1807, in association with the adoption of English as
business language, for trade, in Canton.
Besides, not all creolists agree with the above connection. For example, Alleyne
(1971) and Chaudenson (1989f) argue that creoles (at least those of the Atlantic
and Indian Oceans) started without antecedent pidgins. According to Chaudenson,
plantation communities were preceded by homesteads on which mesolectal approximations of European lexifiers, rather than pidgins, were spoken by earlier slaves (see
also Baker 1982, 1990). Thus, creoles developed gradually as autonomous systems,

Creoles and creolization

moving away from their lexifiers, consistent with the gradual shift from the homestead
kind of dwelling and a farming modus vivendi to the dominant plantation society and
industry. Historical evidence on the peopling of the colonies also suggests that creoles
basilects (the varieties that differ the most from their lexifiers) may have consolidated
their features from individual varieties during the peak growth of plantations (in the
18thcentury for most colonies), when infant mortality was high, when life expectancy was short, when the plantation populations increased by massive importation
of servile labor, and while (as suggested by Baker & Corne 1986) the proportion of
fluent/native speakers of the earlier colonial varieties kept decreasing. It is difficult to
imagine the conditions in these specific colonial contacts which would have favored
the development of pidgins (as different from interlingual transitional varieties pace
Plag 2008, 2009), keeping in mind that linguistically diverse populations were brought
together to interact regularly (i.e. not sporadically) in these new settings.
According to Halls (1962) life-cycle model, as a creole continues to co-exist
with its lexifier, the latter exerts pressure on it to shed some of its creole features.
This developmental hypothesis can in fact be traced back to Schuchardts (1914) explanation of why African-American vernacular English (AAVE) is structurally closer to
North American English than Saramaccan: continuation of contact between AAVE
and White American English and putatively the influence of the latter on the former.
Although they never elaborated on the position, Jespersen (1921) and Bloomfield (1933)
anticipated DeCamp (1971), Bickerton (1973), and Rickford (1987), who invoked
decreolization to account for speech continua in creole communities. DeCamp coined
the term post-creole continuum to account for the direction of accommodations among
speakers in the creole-speaking polity. The variation putatively reflected the gradual
loss of the creole (basilect). However, this conjecture must be interpreted charitably. An
important reason is that if a variety is creole because of the particular sociohistorical
ecology of its genesis, rather than because of its structural peculiarities (Mufwene
2000), it cannot stop being a creole even after some of the features change. Another
reason is that basilectal and mesolectal features continue to co-exist in these communities, suggesting that Creole has not died yet. Lalla & DCosta (1990) present copious
data against decreolization qua debasilectalization in Caribbean English creoles, just
as Mufwene (1994) adduces linguistic and non-linguistic arguments against the same
process in Gullah. Likewise, Poplack & Tagliamonte (1989, 2001) and the publications
in Poplack (2000) argue against the development of AAVE by decreolization, as does
Schneider (1989). On the other hand, Rickford & Handler (1994) show that in the late
18th century Barbados had a creole basilect, which it may now have lost.
Closely related to the above issue is the common assumption that creoles are separate languages, distinct from their lexifiers and the (ex-)colonial varieties thereof spoken
by the descendants of the Europeans. Thus, the nonstandard varieties of French spoken by the descendants of the Europeans in Canada and Louisiana, and on Caribbean
islands such as St. Barth and St. Thomas are not creoles but dialects of French. Likewise,

49

50

Salikoko Mufwene

New World nonstandard varieties of Spanish and Portuguese are not considered creoles,
despite structural similarities which they display with creoles of the same lexifier. In
the absence of structural criteria for defining creoles or for distinguishing dialect from
language, creolists seem to have accepted uncritically the tradition of disfranchising
those varieties of European languages which have been appropriated as vernaculars
by non-European majorities.4 In defense of uniformitarianism in the study of creoles,
DeGraff (2003, 2005, 2009) aptly exposes a long series of double standards that have
been applied especially regarding their developments and their ability to serve the
communicative needs of their speakers.
To date the creoles studied most extensively are those lexified by English and
French. Those of the Atlantic and Indian Ocean are, together with Hawaiian Creole,
the prototypes that have informed most theorizing on creole genesis and creoles
structural features, although, as shown in Mufwene (2004, 2005, 2008), the colonial
settings in which creoles have evolved around the Atlantic and in the Indian Ocean are
quite different from those of Hawaii. The former are associated with the demographically dominant presence of slave labor while the latter were developed by contract
laborers. The slaves were ethnolinguistically mixed, even during the homestead phases
of the colonies, whereas the contract laborers were not.
While the terms creole and creolization have been applied often uncritically to
various contact-induced language varieties, several distinctions which are not clearly
articulated have also been made, for instance, between the above prototypes and
koins, semi-creoles, foreign workers varieties of European languages (e.g. Gastarbeiter
Deutsch), or indigenized varieties of European languages (e.g. Nigerian and Singaporean
Englishes). The nature and importance of all these terms deserve re-examining
(Mufwene 1997, 2005). The notion of semi-creole has been questioned several times,
because no particular set of reliable structural features has been offered to justify the
distinction it suggests. The validity of those proposed by McWhorter (1998, 2001)
has been disputed, most forcefully by DeGraff (2001). Whether a new contact-based
variety is called a creole, a pidgin, or an indigenized variety is largely correlated with
the colonization style, the particular variety that the creators of the new varieties were
exposed to, and the mode of transmission (naturalistic versus scholastic), as explained
in Mufwene (2001).
Among the most central questions of creolistics is that of genesis, which paradoxically has to do less with their origins than with how they developed. What particular
restructuring processes produced creoles? Are there any that are specific to creoles?

. Hjelmslev (1938) and Posner (1985) are exceptional in stating that creoles are dialects of
their lexifiers.

Creoles and creolization

The following hypotheses are the major ones competing today: the substrate, the
superstrate, and the language universal(s) hypotheses.
Substratist positions are historically related to the baby talk hypothesis, which
may be traced back to some French creolists of the 19th century: Bertrand-Bocand
(1849); Baissac (1880); Adam (1883) and Vinson (1882, 1888). Putatively, the substrate
languages previously spoken by the Africans enslaved in the New World and the Indian
Ocean were the primary reason why French and other European lexifiers which they
appropriated as vernaculars were restructured into creoles. In a perspective consistent with Schleichers (1863) linguistic Darwinism, they thought that African languages were primitive, instinctive, in natural state, and simpler than the cultivated
European languages with which they came in contact. Considered less evolved than
Europeans, black Africans were allegedly prevented by their inferior anatomical and
mental features from acquiring the refined phonological and grammatical systems of
the European languages. One had to talk to them as to babies. Creoles systems were
thus considered to be reflections of those of the non-European languages previously
spoken by African slaves, identified as the substratum since Hall (1950).
The substrate influence thesis was revived, without its racist component, by Sylvain
(1936). Although she recognizes influence from Norman and other French dialects,
Sylvain argues that African linguistic influence, especially from the Ewe group of
languages, is very significant.5 She was followed by Turner (1949), who disputed
American dialectologists claim that there was virtually no trace of African languages
in African-American English (AAE), showed phonological and morphosyntactic
similarities between Gullah and some West-African (especially Kwa) languages, and
concluded that Gullah is indebted to African sources (Turner 1949:254). Although
his book is remembered more for its long lists of basket [pet] names related to African
languages, it includes interesting brief discussions of serial verb constructions, nominal
plural, absence of a passive voice, modal use of the preposition fuh [f], and several
other grammatical features which interest creolists nowadays.
Three main schools of the substrate hypothesis have emerged since the 1980s.
The first, which may be associated with, for instance, Alleyne (1980, 1986, 1993, 1996)
and Holm (1986, 1988, 1993) is closer to Turners approach and characterized by what
is also perceived as its main weakness: invocation of influence from diverse African
languages without explaining what kinds of selection principles, if any, account for
this seemingly random invocation of sources. Although most studies in this paradigm

. The position stated here follows from the contents of her book and is contrary to her own final
conclusion (the damning last sentence uncharitably repeated in the literature) that Nous sommes
en prsence dun franais coul dans le moule de la syntaxe africaine ou, comme on classe les langues
daprs leur parent syntaxique, dune langue ewe vocabulaire franais (Sylvain 1936:178).

51

52

Salikoko Mufwene

have been morphosyntactic, one must recognize rare phonological studies such as (in)
Holm (1988, 1993), Carter (1993) and Williams (1993), as well as lexical ones such as
Allsopp (1977), and Wade-Lewis (1988).
The second version of the substrate hypothesis has been dubbed the relexification
hypothesis by its own proponents at the Universit du Quebec Montral who profess
to continue in the footsteps of Sylvain (1936), arguing that Haitian Creole is a French
relexification of languages of the Ewe-Fon (or Fongbe) group. Earlier samples of this
position, now integrated in Lefebre (1998, 2004), may be found in Lefebvre (1986, 1993),
Lefebvre & Lumsden (1989, 1994) and Lumsden (1990). This explanation of creole
genesis has been criticized for several basic shortcomings, including the following:
(1)its comparative approach has not taken into account several features which Haitian
Creole shares with nonstandard varieties of French (even if it shares them with Ewe-Fon
languages); (2) it downplays features which Haitian Creole shares also with several
other African languages which were represented in Haiti during the critical stages
of its development; (3) it does not account for those cases where Haitian Creole has
selected structural options which are not consistent with those of Ewe-Fon languages,
nor does it account for cases where similarities with Ewe-Fon languages are only partial.
Moreover, relexificationists assume that languages of the Ewe-Fon group are structurally identical (if not mutually intelligible) and that no competition of influence was
involved among them. This assumption is disputable. The most informative critiques of
the relexification hypothesis are provided by DeGraff (2002) and Aboh (2006, 2009).
The least disputed version of the substrate hypothesis is that presented in Keesing
(1988), which shows that substrate languages may impose their structural features on
the new, contact-induced varieties if they are typologically largely homogeneous, with
most of them sharing the relevant features. Thus Melanesian pidgins are like (most of)
their substrates in having DUAL/PLURAL and INCLUSIVE/EXCLUSIVE distinctions
and in having a transitive marker on the verb. Sankoff & Brown (1976) had shown
similar influence with the bracketing of relative clauses with ia. Sankoff (1993) does
the same in highlighting substrate influence on the focus construction. However, the
pidgins have not inherited all the peculiarities of Melanesian languages. For instance,
they do not have their VSO major constituent order, nor do they have much of a
numeral classifying system in the combination of pela with quantifiers.
Competing with the above genetic views has been the dialectologist, or superstrate hypothesis, according to which the primary, if not the exclusive, sources of creoles structural features are nonstandard varieties of their lexifiers. Speaking of AAE,
Krapp (1924); Kurath (1928); Johnson (1930), and Crum (1940) claimed that this variety
was an archaic retention of the nonstandard speech of low-class whites with whom
the African slaves had been in contact. According to them, African substrate influence was marginal, in the form of some isolated lexical items such as goober peanut,
gumbo, and okra. It would take until McDavid (1950) and McDavid & McDavid (1951)

Creoles and creolization

before some African grammatical contributions to AAE were recognized. Up to the

1980s, dialectologists treated African influence on AAVE as marginal. DEloia (1973)
and Schneider (1981, 1982) invoke several dialectal English models to rebut Dillards
(1972) thesis that AAVE developed from an erstwhile West-African Pidgin English
brought over by slaves. Schneider (1993) is more conciliatory.
Regarding French creoles, the dialectologist position was first defended by Faine
(1936, 1939), according to whom Haitian Creole was essentially Norman French. This
position was espoused later by Hall (1958), who argues that the basic relationship of
Creole is with 17th-century French, with heavy carry-overs or survivals of African linguistic structure (on a more superficial structural level) from the previous language(s)
of the earliest speakers of Negro Pidgin French; that its lexical relationship is with
nineteenth- and twentieth-century French (1958: 372). Goodman (1964); Chaudenson
(1973) and Posner (1985) see things in basically the same way. Chaudenson (1989,
1992, 2001, 2003), however, has been more accommodating to substrate influence,
arguing that structures of creoles would be more similar to those of their non-creole
ex-colonial kin (e.g. Qubcois and the French vernacular of St. Barth), had it not been
for the influence of the substrate languages on colonial French.
The universalist hypotheses, which emerged as strong contenders in 1980s and
1990s, have their forerunners in the 19th century. Adolfo Coelho (18801886) partly
anticipated Bickertons (1981, 1984) language bioprogram hypothesis in stating that
creoles owe their origin to the operation of psychological or physiological laws that
are everywhere the same, and not to the influence of the former languages of the
people among whom these dialects are found (1880: 193,195). Bickerton pushed
things further in claiming that children made creoles by fixing the parameters of
these new language varieties in their unmarked, or default, settings as specified in
Universal Grammar. According to Bickerton (1986, 1992) this creole grammar is
determined by the vocabulary acquired from the pre-existing, syntax-less pidgin of
the parent generation. To account for cross-creole structural differences, Bickerton
(1984) invoked a Pidginization Index factor, whose explanatory capacity was disputed
in Singler (1990).
Some nagging questions about Bickertons position include the following: How
can we explain similarities between the creoles, putatively formed by children, and
expanded pidgins (e.g. Cameroon Pidgin English), whose structural expansion and
vernacularization need not be associated with more extensive usage by children?
Is there convincing evidence for assuming that adult speech is less controlled by
Universal Grammar than child language is? How can we account for similarities
between the allegedly abrupt creolization process and second-language acquisition?
If syntax is determined by the lexicon, does it make sense to assume default parametric
settings at all in Universal Grammar? Is the hypothesis consistent with the social
histories of the territories where prototypical creoles have developed?

53

54

Salikoko Mufwene

Not all creolists who have invoked universalist explanations one way or another
have made children critical to the emergence of creoles. For instance, Sankoff (1979) and
Mhlhusler (1981, 1983) make allowance for this device to operate in adults, too.
Nowadays, few creolists subscribe to one exclusive genetic account, as evidenced
by, e.g. Mufwene, ed. (1993). The complementary hypothesis (Baker & Corne 1986;
Baker 1993; Hancock 1986, 1993 & Mufwene 1986, 1993) seems to be an adequate
alternative, provided we can articulate the conditions (linguistic and nonlinguistic)
under which the competing influences (between the substrate and superstrate languages,
and within each group) may converge or prevail upon each other. The specific ways in
which the competition and selection mechanisms that are involved produce this evolution process have been elaborated in Mufwene (2001, 2002, 2005) and Aboh (2006,
2009). Works such as Arends (1987, 1995), Baker (1982, 1993), Chaudenson (1992,
2001, 2003), Corne (1999); Singler (1993); Siegel (2000) and Roberts (2005) have
contributed a great deal of sociohistorical information that informs the approach and
the hypothesis of a gradual, rather than abrupt, evolution for creoles.
Still, the future of research on the development of creoles has some problems to
overcome. So far information on the nonstandard varieties of the lexifiers spoken by
the European colonists remains limited. There are still few comprehensive descriptions of creoles structures, which makes it difficult to determine globally how the
competing influences interacted among themselves and how the features selected from
diverse sources became integrated into new systems. Very few structural facts have
been correlated with the conclusions suggested by the sociohistorical backgrounds
of individual creoles. Other issues remain up in the air, for instance, regarding the
markedness model that is the most adequate to account for the selection of features
into creoles systems and for an adequate interpretation of substrate influence and of
the bioprogram qua Universal Grammar.6
There is much more literature on the development, sociology, and morphosyntax
of these language varieties than on their phonologies, semantics, and pragmatics.
With the exception of time reference (e.g. Singler 1990) and nominal number (see
Tagliamonte& Poplack 1994 for references), studies in the latter two subfields are

. For developmental issues on pidgins and creoles, the following books are good starting
points: Hymes (1971), Valdman (1977), Hill (1979), Muysken & Smith (1986), Mufwene (1993),
Ansaldo et al. 2007, Siegel 2008, Aboh & Smith 2009, and Kouwenberg & Singler (2008). More
general issues may be checked in volumes of the Creole Language Library series and of Amsterdam
Creole Studies, in the Journal of Pidgin and Creole Languages, and in Etudes Croles. The latter
journal features mostly articles on developmental, anthropological, and literary aspects of French
creoles, but several issues since the mid-1980s include some on non-French creoles. Several issues
of Pacific Linguistics also include publications on Melanesian creoles.

Creoles and creolization

scant. On the other hand, the development of quantitative sociolinguistics owes a lot to
research on AAVE since the mid-1960s (see, e.g. Labov 1972) and Caribbean English
creoles (e.g. Rickford 1987). Numerous publications in American Speech, Language in
Society, and Language Variation and Change reflect this.7
Studies of structural aspects of creoles have yet to inform general linguistics beyond
the subject matter of serial verb constructions and, to some extent, time reference.
For instance, studies of lectal continua have had this potential, but little has been done
by creolists to show how their findings (can) apply to other languages. The mixed
nature of mesolects, those intermediate varieties combining features of both the
acrolect (the educated variety of the lexifier) and the basilect (the opposite creole
extreme) should have informed general linguistics against the fallacy of assuming
monolithic grammatical systems (Mufwene 1992b). However, little has been done
on the subject matter. Likewise, the debate on creole genesis could have informed
historical linguistics on the importance of varying external conditions to language
change (Mufwene 2001, 2005).
Although lack of consensus among creolists may be invoked as a general reason
for this failure to influence general linguistics, alarming indifference from theoretical
linguists, especially those engaged in theories of typology and universals, is a more
important reason. Consensus cannot be expected in creolistics any more than in
other subfields of linguistics. However, in the broader context of language contact
(including second-language acquisition), studies of creole genesis especially have been
inspiring. For instance, Thomason & Kaufman (1988) is widely cited in studies of indigenized Englishes. Andersen (1983), whose example is well followed by Kouwenberg&
Patrick (2003), was an important step toward consolidating common interests between
second-language acquisition and creole genesis. More cross-fertilization might be
expected between studies of creole genesis and those of (child) language development
(DeGraff 1996), as among diverse subfields of linguistics.

References
Aboh, E.O. (2006). The role of the syntax-semantics interface in language transfer. In C. Lefebvre,
L.White & C. Jourdan (eds.) L2 acquisition and creole genesis: 253275. Benjamins.
Aboh, E.O. (2009). Competition and selection: Thats all! In Aboh & Smith (eds): 317344.
Aboh, E.O. & N. Smith (Eds.) (in press). Complex processes in new languages. Benjamins.
Adam, L. (1883). Les idiomes ngro-aryens et malayo-aryens. Maisonneuve.

. There are also several surveys of creolistics today, including the following: Mhlhusler (1986);
Romaine (1988); Holm (1988), Arends, Muysken & Smith (1994), Manessy (1995) and Siegel
(2008). They vary in geographical areas of focus and adequacy.

55

56

Salikoko Mufwene
Alleyne, M.C. (1971). Acculturation and the cultural matrix of creolization. In D.H. Hymes (ed.):
169186.
(1980). Comparative Afro-American. Karoma.
(1986). Substratum influences. In P. Muysken & N. Smith (eds.): 301315.
(1993). Continuity versus creativity in Afro-American language and culture. In S. Mufwene
(ed.): 167181.
(1996). Syntaxe historique crole. Karthala.
Allsopp, R. (1977). Africanisms in the idiom of Caribbean English. In P.F. Kotey & H. Der-Houssikian
(eds.) Language and language problems in Africa: 135142. Hornbeam Press.
Andersen, R. (ed.) (1983). Pidginization and creolization as language acquisition. Newbury House.
Ansaldo, U., S. Matthews & L. Lim (Eds.) (2007). Deconstructing Creole. Benjamins.
Arends, J. (1987). Syntactic developments in Sranan. PhD. Diss., University of Nijmegen.
(ed.) (1995). The early stages of creolization. John Benjamins.
Arends, J., P. Muysken & N. Smith (eds.) (1994). Pidgins and creoles. John Benjamins.
Arveiller, R. (1963). Contribution ltude des termes de voyage en franais (15051722). DArtrey.
Baissac, C. (1880). Etude sur le patois crole mauricien. Berger-Levrault.
Baker, P. (1982). The contribution of non-Francophone immigrants to the lexicon of Mauritian Creole.
PhD. Diss., SOAS, University of London.
(1990). Off target? Journal of Pidgin and Creole Languages 5: 107119.
(1993). Assessing the African contribution to French creoles. In S. Mufwene (ed.): 123155.
Baker, P. & C. Corne (1986). Universals, substrata and the Indian Ocean creoles. In P. Muysken &
N.Smith (eds.): 163183.
Baker, P. & P. Mhlhusler (1990). From business to pidgin. Journal of Asian Pacific Communication
1: 87115.
Bertrand-Bocand, E. (1849). Notes sur la Guine portugaise ou Sngambie mridionale. Bulletin
de la Socit de Gographie 12: 5793.
Bickerton, D. (1973). The nature of a creole continuum. Language 49: 640669.
(1981). Roots of language. Karoma.
(1984). The language bioprogram hypothesis. Behavioral and Brain Sciences 7: 173221.
(1986). Beyond roots. Journal of Pidgin and Creole Languages 1: 225232.
(1992). The creole key to the black box of language. In M. Ptz (ed.) Thirty years of linguistic
evolution: 97108. John Benjamins.
(1999). How to acquire language without positive evidence: What acquisitionists can learn from
creoles. In M. DeGraff (ed.) Language creation and language change: Creolization, diachrony,
and development: 4974. MIT Press.
Bloomfield, L. (1933). Language. Holt, Rinehart & Winston.
Carter, H. (1993). Vowel length in Afro-American. In S. Mufwene (ed.): 328345.
Chaudenson, R. (1973). Pour une tude compare des croles et parlers franais doutre-mer. Revue
de Linguistique Romane 37: 342371.
(1986). And they had to speak any way . In J. Fishman (ed.) The Fergusonian impact, vol. 1:
6982. Mouton de Gruyter.
(1989). Croles et enseignement du franais. LHarmattan.
(1992). Des les, des hommes, des langues. LHarmattan.
(2001). Creolization of language and culture. Routledge.
(2003). Croles, thorie, applications, implications. LHarmattan.
Coelho, F.A. (18801886). Os dialectos romnicos ou neolatinos na Africa, Asia, ae America. Bolletim da
Sociedade de Geografia de Lisboa 2: 129196 (18801881); 3: 451478 (1882); 6: 705755 (1886).

Creoles and creolization

Corne, C. (1999). From French to Creole: The development of new vernaculars in the French colonial
world. University of Westminster Press.
Crum, M. (1940). Gullah. Duke University Press.
Decamp, D. (1971). Toward a generative analysis of a post-creole speech continuum. In D.H. Hymes
(ed.): 349370.
DeGraff, M. (2001). On the origin of creoles: A Cartesian critique of Neo-Darwinian linguistics.
Linguistic Typology 5: 213310.
(2002). Relexification: A reevaluation. Anthropological Linguistics 44: 321414.
(2003). Against creole exceptionalism. Discussion note. Language 79: 391410.
(2005). Linguists most dangerous myth: The fallacy of creole exceptionalism. Language in
Society 34: 533591.
(2009). Language acquisition in creolization and, thus, language change: Some Cartesianuniformitarian boundary conditions. Language and Linguistics Compass 3/4: 888971. <http://
www3.interscience.wiley.com/journal/122464861/abstract>.
(ed.) (1996). Creolization, diachrony, and language acquisition. MIT Press.
DEloia, S.G. (1973). Issues in the analysis of Negro nonstandard English. Review of Dillard (1972).
Journal of English Linguistics 7: 87106.
Dillard, J.L. (1972). Black English. Random House.
Faine, J. (1936). Philologie crole: tudes historiques et tymologiques sur la langue crole dHaiti.
Imprimerie de lEtat.
(1939). Le crole dans lunivers. Tome 1: Le mauricien. Imprimerie de lEtat.
de Fral, C. (1989). Pidgin-English du Cameroun. Peters/SELAF.
Goodman, M. (1964). A comparative study of French Creole dialects. Mouton.
Hall, R.A., Jr. (1950). The genetic relationships of Haitian Creole. Ricerche Linguistiche 1: 194203.
(1958). Creole languages and genetic relationships. Word 14: 367373.
(1962). The life-cycle of pidgin languages. Lingua 11: 151156.
(1966). Pidgin and creole languages. Cornell University Press.
Hancock, I. (1986). The domestic hypothesis, diffusion and componentiality. In P. Muysken &
N.Smith (eds.): 71102.
(1993). Creole language provenance and the African component. In S. Mufwene (ed.): 182191.
Hill, K.C. (ed.) (1979). The genesis of language. Karoma.
Hjelmslev, L. (1938). Etudes sur la notion de parent linguistique. Premire tude. Revue des Etudes
Indo-Europennes 1: 271286.
Holm, J. (1986). Substrate diffusion. In P. Muysken & N. Smith (eds.): 259278.
(1988). Pidgins and creoles. Vol. 1. Cambridge University Press.
(1989). Pidgins and creoles. Vol. 2. Cambridge University Press.
(1993). Phonological features common to some West-African and Atlantic creole languages.
In S. Mufwene (ed.): 317327.
Hymes, D.H. (ed.) (1971). Pidginization and creolization of languages. Cambridge University Press.
Jespersen, O. (1921). Language. W.W. Norton & Co.
Johnson, G. (1930). Folk culture on St. Helena Island, South Carolina. University of North Carolina
Press.
Keesing, R.M. (1988). Melanesian Pidgin and the Oceanic substrate. Stanford University Press.
Kouwenberg, S. & P. Patrick (eds.) (2003). Special issue on Reconsidering the role of SLA in pidginization
and creolization. Studies in Second Language Acquisition 25(2).
Kouwenberg, S. & J.V. Singler (eds.) (2006). The Handbook of Pidgin and Creole Studies. Blackwell.
Krapp, G.P. (1924). The English of the Negro. The American Mercury 2: 190195.

57

58

Salikoko Mufwene
Kurath, H. (1928). The origin of dialectal differences in spoken American English. Modern Philology.
25: 385395.
Labov, W. (1972). Language in the inner city. University of Pennsylvania Press.
Lalla, B. & J. DCosta (1990). Language in exile. University of Alabama Press.
Lefebvre, C. (1986). Relexification in creole genesis revisited. In P. Muysken & N.Smith (eds.): 279300.
(1993). The role of relexification and syntactic analysis in Haitian Creole. In S. Mufwene (ed.):
254279.
(1998). Creole genesis and the acquisition of grammar: The case of Haitian Creole. Cambridge
University Press.
(2004). Issues in the Study of Pidgin and Creole Languages. John Benjamins.
Lefebvre, C. & J.S. Lumsden (1994). Le rle central de la relexification dans la gense des croles.
Pluringuismes 8: 4782.
(eds.) (1989). Special issue on La crolisation. The Canadian Journal of Linguistics 34: 247376.
Lumsden, J.S. (1990). The biclausal structure of Haitian clefts. Linguistics 28: 741759.
Manessy, G. (1995). Croles, pidgins, varits vhiculaires. CNRS Editions.
Markey, T.L. (1982). Afrikaans: Creole or non-creole? Zeitschrift fur Dialektologie und Linguistik
2:169207.
McDavid, R., Jr. (1950). Review of Lorenzo Dow Turners Africanisms in the Gullah dialect. Language
26: 323333.
McDavid Jr., R. & V. McDavid (1951). The relationship of the speech of the American Negroes to the
speech of whites. American Speech 26: 317.
McWhorter, J.H. (1998). Identifying the creole prototype: Vindicating a typological class. Language
74: 788818.
(2001). The worlds simplest grammars are creole grammars. Linguistic Typology 5: 125166.
Mufwene, S.S. (1986a). Les langues croles peuvent-elles tre dfinies sans allusion leur histoire?
Etudes Croles 9: 135150.
(1986b). The universalist and substrate hypotheses complement one another. In P. Muysken&
N. Smith (eds.): 129162.
(1988). Why study pidgins and creoles? Journal of Pidgin and Creole Languages 3: 265276.
(1992a). Africanisms in Gullah. In J. Hall, N. Doane & D. Ringler (eds.) Old English and new:
156182. Garland.
(1992b). Why grammars are not monolithic. In D. Brentari, G. Larson & L.A. MacLeod (eds.)
The joy of grammar: A festschrift in honor of James D. McCawley: 225250. Benjamins.
(1993). African substratum. In S. Mufwene (ed.): 192208.
(1994). On decreolization. In M. Morgan (ed.) Language, loyalty, and identity in creole situations:
6399. Center for Afro-American Studies.
(1995). Jargons, pidgins, creoles, and koins. In A. Spears & D. Winford (eds.) Pidgins and
creoles. John Benjamins.
(1997). Jargons, pidgins, creoles, and koins: What are they? In A. Spears & D. Winford (eds.)
The structure and status of pidgins and creoles: 3570. John Benjamins.
(2000). Creolization is a social, not a structural, process. In I. Neumann-Holzschuh &
E.Schneider (eds.) Degrees of restructuring in creole languages: 6584. John Benjamins.
(2001). The ecology of language evolution. Cambridge University Press.
(2002). Competition and selection in language evolution. Selection 3: 4556.
(2004). Multilingualism in linguistic history: Creolization and indigenization. In T. Bhatia &
W. Ritchie (eds.) Handbook of bilingualism: 460488. Blackwell.
(2005). Croles, cologie sociale, volution linguistique. LHarmattan.
(2008). Language evolution: contact, competition and change. Continuum Press.

Creoles and creolization

(ed.) (1993). Africanisms in Afro-American language varieties. University of Georgia Press.

Mhlhusler, P. (1981). The development of the category of number in Tok Pisin. In P. Muysken (ed.)
Generative studies on creole languages: 3584. Foris.
(1983). The development of word-formation in Tok Pisin. Folia Linguistica 17: 463487.
(1985). The number of pidgin Englishes in the Pacific. Papers in Pidgin and Creole Linguistics
No. 1 Pacific Linguistics, A-72: 2551.
(1986). Pidgin and creole linguistics. Blackwell.
Muysken, P. & N. Smith (eds.) (1986). Substrata versus universals in creole genesis. John Benjamins.
Oldendorp, C.G.A. (1777). Geschichte der Mission der evangelischen Brueder auf den caraibischen
Inseln S. Thomas, S. Croix und S. Jan. Christian Friedrich Laur.
Plag, I. (2008). Creoles as interlanguages: Syntactic structures. Column. Journal of Pidgin and Creole
Languages 23: 307328.
( 2009). Creoles as interlanguages: Phonology. Column. Journal of Pidgin and Creole
Languages 24: 119138.
Poplack, S. (2000). The English history of African American English. Blackwell.
Poplack, S. & S. Tagliamonte (1989). Theres no tense like the tresent: Verbal -s inflection in early
Black English. Language Variation and Change 1: 4784.
(1991). African American English in the diaspora. Language Variation and Change 3: 301339.
(1994). -S or nothing. American Speech 69: 227259.
(2001). African American English in the Diaspora. Blackwell.
Posner, R. (1985). Creolization as typological change. Diachronica 2: 167188.
Rickford, J.R. (1987). Dimensions of a creole continuum. Stanford University Press.
Rickford, J.R. & J.S. Handler (1994). Textual evidence on the nature of early Barbadian speech,
16761835. Journal of Pidgin and Creole Languages 9: 221255.
Roberts, S.J. (1998). The role of diffusion in the genesis of Hawaiian Creole. Language 74: 139.
(2005). The emergence of Hawaii Creole English in the early 20th century: The sociohistorical
context of creole genesis. Ph.D. dissertation. Stanford University.
Romaine, S. (1988). Pidgin and creole languages. Longman.
Sankoff, G. (1979). The genesis of a language. In K.C. Hill (ed.): 2347.
(1993). Focus in Tok Pisin. In F. Byrne & D. Winford (eds.) Focus and grammatical relations
in creole languages: 117140. John Benjamins.
Sankoff, G. & P. Brown (1976). The origins of syntax in discourse. Language 52: 631666.
Schleicher, A. (1863). Die darwinsche Theorie und die Sprachwissenschaft. Bhlau.
Schneider, E.W. (1981). Morphologische und syntaktische Variablen im amerikanischen Early Black
English. Lang.
(1982). On the history of Black English in the USA. English World-Wide 3: 1846.
(1989). American earlier Black English. University of Alabama Press.
(1993). Africanisms in the grammar of Afro-American English. In S. Mufwene (ed.): 208221.
Schuchardt, H. (1909). Die Lingua Franca. Zeitschrift fur Romanische Philologie 33: 441461.
(1914). Die Sprache der Saramakkaneger in Surinam. Muller.
Siegel, J. (2000). Substrate influence in Hawaii Creole English. Language in Society 29: 197236.
(2008). The emergence of pidgin and creole languages. Oxford University Press.
Singler, J.V. (1990). On the use of sociohistorical criteria in the comparison of creoles. Linguistics
28:645659.
(1993). African influence upon Afro-American language varieties. In S. Mufwene (ed.):
235253.
(ed.) (1990). Pidgin and creole tense-mood-aspect systems. John Benjamins.
Sylvain, S. (1936). Le crole hatien. De Meester.

59

60

Salikoko Mufwene
Tagliamonte, S. & S. Poplack (1988). Tense and aspect in Saman English. Language in Society
17:513533.
(1993). The zero-marked verb. Journal of Pidgin and Creole Languages 8: 171206.
Thomason, S.G. & T. Kaufman (1988). Language contact, creolization, and genetic linguistics. University
of California Press.
Turner, L.D. (1949). Africanisms in the Gullah dialect. University of Chicago Press.
Valdman, A. (ed.) (1977). Pidgin and creole linguistics. Indiana University Press.
Valkhoff, M.F. (1966). Studies in Portuguese and creole with special reference to South Africa.
Witwatersrand University Press.
Vinson, J. (1882). Crole. Dictionnaire des sciences anthropologiques et ethnologiques. Paris.
(1888). La linguistique. In Berthelot (ed.) La grande encyclopdie, vol. 22: 286296. Paris.
Wade-Lewis, M. (1988). The African substratum in American English. PhD. Diss., New York University.
Williams, S.W. (1993). Substantive Africanisms at the end of the African linguistic diaspora. In
S.Mufwene (ed.): 406422.

Dialect
Ronald Macaulay

1. Introduction
The term dialect has generally had a bad press. Bloomfield, for example, noted that local
dialects are spoken by the peasants and the poorest people of the towns (1933:50)
though he also thought that the lower middle class spoke sub-standard speech. More
than 60 years later Hudson observes that the contrast between language and dialect
is a question of prestige, a language having prestige which a dialect lacks (1996:36).
(He goes on to make the interesting claim that Standard English is consequently a
language, which would imply that nonstandard dialects are not part of the same language.) It is hardly surprising that in the more enlightened times of the 1960s linguists
tended to shy away from referring to dialects and looked for a more neutral term that
would have less negative connotations (see Section2). There was also an unfortunate
period in the development of generative grammar in the U.S., when disagreements
about subjective judgments regarding the grammaticality of dubious sentences were
labeled dialect differences. Such extravagant usage was enough to bring any term into
disrepute. However, in recent years many linguists have reverted to the use of the term
dialect when dealing with linguistic variation, though the precise definition of the
term remains unclear.
For example, McDavid (1979:70) says that a dialect is simply a habitual variety
of a language, regional or social, which leaves its nature and scope undefined. Francis
admits the difficulty in drawing boundaries but goes on to say: It is obvious that
there are dialects; that people in, for example, Charleston, South Carolina, use different words, different pronunciation, and different grammar from people in Boston or
New York (Francis 1983:6). Chambers & Trudgill (1980:5) restrict the term dialect
to varieties that are grammatically (and perhaps lexically) distinct as opposed to differences in pronunciation which they label accent, whereas Labov (1991) bases his
identification of dialects in the U.S. solely on pronunciation. The use of the term dialect
is a good example of Saussures observation that in the matter of language, people have
always been satisfied with ill-defined units (1966:111).
There are two other problems with the notion of dialect. Haugen (1966) points
out that the relationship between a language and a dialect is a superordinate one.
IfX is a dialect of Y, then Y cannot be a dialect of X. The problem with viewing dialects
in this perspective is that one becomes involved in an endless regression leading back
to the original ur-language, so that every form of speech is, in this sense, a dialect.
This poses no problem for the definition of languages because, as Haugen points out,

62

Ronald Macaulay

it is nations that have languages, with standards enforced not so much nowadays by
an army and a navy (as has frequently been claimed) but mainly through the educational system, the mass media, and the central bureaucracy (Cameron 1995). However, defining the superordinate form does not necessarily help with delimiting the
subordinate units.
Another way of looking at the relationship of a language and its dialects is that
of heteronomy:
A linguistic system will be heteronomous in terms of another, historically related one
when the former functions in the linguistic community as a dependent variety of
the latter, and is consequently subject to correction in its direction, i.e. is subject to
regular structural readjustment so that it will come to resemble the other more closely.
(Stewart 1968:535)

However, it is not the case that all dialects are in a process of readjustment in the
direction of the superordinate language. There are also what have been called roofless dialects (Kloss 1978; Cadiot 1987), such as Middle English, Scots, Low German,
and Occitan. The speakers of such dialects do not have an obvious superordinate
language. Most speakers of these dialects will also consider themselves to be speakers
of some superordinate language (e.g. French in the case of Occitan speakers) but
need not feel pressured to modify their other form of speech in the direction of the
superordinate variety.
The other perennial problem with the definition of the term dialect is the dialect
continuum or what may be called the Gaston Paris phenomenon. Paris pointed out that
a traveler passing through contiguous areas where the inhabitants have no difficulty
understanding their neighbors will reach a point where the form of speech would no
longer be comprehensible to the people in the district where he had begun. Paris concluded that there were no dialects because the forms of speech in one location blended
imperceptibly into those of the next community (Gauchat 1903). The permeability of
dialect boundaries has proved a perennial problem for dialectologists.

2. Other labels
2.1 Variety
The favorite alternative label for dialect was for some time variety:
We shall use variety as a neutral term to apply to any particular kind of language
which we wish, for some purpose, to consider as a single entity. The term will be used
in an ad hoc manner in order to be as specific as we wish for a particular purpose.
We can, for example, refer to the variety Yorkshire English, but we can equally well
refer to Leeds English as a variety, or middle-class Leeds English and so on.
(Chambers& Trudgill 1980:5)

Dialect

However, since variety is endless, perhaps the most useful function of this term has
been mainly to put lay persons in their place when they make disparaging remarks
about dialects. It is still used as a non-committal term when the status of the form of
speech is unclear, but it is not a term used with a precise definition and it does not contribute to any theory of language. Its use is noncontroversial for the very good reason
that it does not make any theoretical claims about the status of the entity referred to
in this way.
2.2 Lect
Another term that was introduced in the 1970s was lect for any combination of
linguistic differences (Bailey 1973:162), which had a certain popularity, possibly
following on the multiple layers postulated to exist in a creole continuum between the
basilect (the form of speech with the lowest prestige in the whole community) and the
acrolect (the form with the highest prestige). Bickerton (1973:643) argued for the use
of the term lect in a dynamic paradigm to account for linguistic variation:
to speak of dialects or even perhaps languages may be misleading; these terms
merely seek to freeze at an arbitrary moment, and to coalesce into an arbitrary whole,
phenomena which in nature are on-going and heterogeneous.

However, it was never clear what kind of existence lects were supposed to have or how
they were to remain liquid (while all around were freezing).
The most interesting use of the term was perhaps that by Chambers & Trudgill
(1980) in their description of mixed lects and fudged lects. They illustrated the
phenomenon with reference to the development of Middle English short u. South of
a line running roughly from the Severn to the Wash the vowel is rounded but north
of this line the vowel remains unrounded. Chambers & Trudgill showed that some of
the Survey of English Dialect respondents near this boundary had mixed lects with
rounded vowels in some words and unrounded vowels in others. They also noticed
that some of the respondents had an intermediate vowel in some words and they
considered these speakers to have fudged lects. However, it is not obvious that their
account would have been any different if they had referred to mixed/fudged dialects.
Since then the use of the term lect seems to have receded.
2.3 Vernacular
A more complex notion is that of the vernacular, which is particularly problematic.
Labov introduced the term into sociolinguistics when he coined the expression Black
English Vernacular (BEV) to refer to that relatively uniform grammar found in its most
consistent form in the speech of Black youth from 8 to 19 years old who participate
fully in the street culture of the inner cities (1972:xiii). Prior to this time, the most
common use of the term was to refer to what Haugen (1966:929) called undeveloped

63

64

Ronald Macaulay

languages that might later develop into national standardized languages, as happened,
for example, with the present national languages of Europe. Following Labovs example,
other sociolinguists extended the use of vernacular more generally to that of a highly
focused variety at the lowest stratum (Milroy 1980:180). Labov (1981) further complicated the use of the term by employing it both with reference to community norms
and also to the basic form of speech used by an individual. As a result, there has been
a tendency to privilege a particular form of speech that I have called the most casual
style of the basilect (Macaulay 1997a:16). While this is a perfectly legitimate object of
research it should not be confused with the investigation of dialects.
2.4 Social dialects
Sociolinguists have further complicated the terminological situation by talking about
social dialects (or sometimes sociolects). This may seem a reasonable extension of the
term. If variation on the horizontal plane of geographical distance is dialectal, then
why not treat the vertical dimension of socio-economic differences as a similar pheno
menon? But this is the camels nose under the tent-flap. In addition to social class differences, there are also gender, age, occupational, and ethnic differences. These will be
found in most local dialects and the differences will contribute to the features available
to the speakers of that dialect; they are not separate from the dialect. Moreover, if we
begin to look at the totality of language variation, we run up against other ill-defined
notions such as register and style which will be part of the repertoire of the speakers but
again not separate from the dialect. To claim that there are social dialects would appear
to suggest that local dialects are not social, and that would be wrong.

3. Dialects as local forms of speech

More than forty years ago Andr Martinet remarked:
[The dialectologists] were, and often still are, antiquarians looking for rare survivals
rather than linguists interested in characterizing as a whole the form of speech they
are faced with. This may explain why dialectologists, the name of their avowed
profession notwithstanding, have not presented us with a clear definition and semantic
delimitation of the term dialect. (Martinet 1954:2)

Present-day dialectologists and sociolinguists are not usually looking for rare survivals
but neither have they always seemed to be interested in characterizing as a whole the
form of speech they are faced with. Instead of rare survivals, present-day dialecto
logists and sociolinguists have often focused on linguistic variables, a term introduced
by Labov. However, this approach will not necessarily help us to meet Martinets goal

Dialect

of characterizing any form of speech as a whole, because some of the distinguishing

features of a dialect may not be variables of this kind.
It is perhaps not surprising that dialectologists have not been quick to make strong
claims about dialect boundaries.1 As Kretschmar & Schneider (1996:2526) point out,
if one defines the boundaries of an area it is possible to describe all available data and
call them a list of dialect features but the boundaries may be arbitrary and the dialect
may have no essence, no reality apart from having been bounded and described. On
the other hand, if boundaries must emerge from the evidence, nobody knows how
many features are required to establish the existence of a dialect.
There is a fundamental problem with the latter goal of attempting to identify dialects on the basis of the distribution of variants. The number of groups that could be
identified on the basis of shared linguistic characteristics is impossibly large (Macaulay
1997:117) and there is no guarantee that any particular one of these groupings will be
of general interest unless there are both social and linguistically interesting reasons for
the group to share certain characteristics. For example, the number of people who are
familiar with the register of games such as chess or bridge is, in principle, identifiable,
but they would probably not form a particularly interesting linguistically determined
group, nor would those familiar with the register of, say, baseball, gourmet cooking,
or medical terms (Wolfram & Cavendar 1992). Attempting to identify dialects solely
on the basis of shared forms is doomed to failure because both communities and their
forms of speech are heterogeneous.
In recent years linguists have shown greater interest in the notion of dialects.
Trudgill (1986), for example, in a book entitled Dialects in Contact deals with a range
of dialects, from such large scale notions as English English, American English,
Australian English, and other colonial varieties, to the language of communities as
small as minor towns in East Anglia or Scandinavia. Does it make sense to consider
such diverse phenomena under the single label of dialects? To my mind, yes. If the
term dialect is to have any coherent interpretation it must refer to a form of speech
based on geographical location, preferably one that the speakers themselves would
most likely admit to using. Most speakers of American English or Australian English

. For many reasons, there are going to be fuzzy edges to any dialect survey. In most cases it will
not make sense to draw precise lines and say On this side X and On that side Y. In many cases
it will not be the presence or absence of a form that will be significant but its relative frequency.
I illustrated this point with examples from the Linguistic Survey of Scotland and the Survey of
English Dialects (Macaulay 1985). I was able to show how the results from the surveys could be
better displayed by representing the percentage (i.e. frequency) of response for a particular form
within a given area rather than treating single occurrences as significant. In one case this served to
identify a relic area for a recessive form, a fact which could not be gleaned from the corresponding
maps in The Linguistic Atlas of England (Orton, Sanderson & Widdowson 1978).

65

66

Ronald Macaulay

will not expect to be identified with speakers of English English, regardless of the
extent to which they believe that the latter has prestige. Similarly, most of the inhabi
tants of Glasgow or Belfast will not be surprised to be recognized as such by their
form of speech, despite the low esteem in which both accents are held. The goal of
dialectology should be to tabulate those features of speech that characterize a particular
locality regardless of its size.
Where the traditional dialectologists ran into trouble was in their apparent desire
to describe a pure dialect. There can, of course, be no such thing, and this is probably
what Gaston Paris meant when he said that there are no dialects. In any community
there will be differences related to social status, education, gender, age, and, in many
cases, ethnicity. These differences can be studied for their own sake or as part of a
larger survey of the speech community. However, this does not mean that the notion
of a local form of speech is meaningless. What identifies a locality as a dialect area will
depend, as Kurath observes, on a combination of socio-cultural factors (1972:76)
such as political domains or settlement areas.2
Restricting the term dialect to refer to a local form of speech might have helped
in the confused debate that followed the December 1996 decision of the Oakland
School Board in California to recognize Ebonics (i.e. African American Vernacular
English) as a separate language. This claim was met with outrage and derision by the
general public, many people accusing the school board of a cynical attempt to obtain
funds available for bilingual education. The efforts of some linguists to clarify the situation were not helped by their own doubts as to whether African American Vernacular
English (AAVE) is a dialect of English or not. They were attempting to counter the
negative view of AAVE as slang, street talk, or simply bad English. If AAVE is a
dialect of English, then how does it differ from other dialects of English which do not
require special treatment? If it is not a dialect, what is it? My own view is that African
American English (AAE) is a dialect of English in the same sense that American
English or Canadian English is a dialect, but African American English is not a dialect

. An attempt to investigate a linguistic area on the basis of traditional dialect survey materials
was my examination of the lexical items collected in the Linguistic Survey of Scotland (Mather &
Speitel 1975,1977) in the county of Ayr (Macaulay 1997b). I looked at those items that were candidates for being typical Ayrshire words either because they were given by a large number of Ayrshire
respondents or because they were given by few respondents outside of Ayrshire. The results did not
provide an overwhelming argument for Ayrshire as a unique dialect distinct from its neighbors.
Disappointing as it might seem to those who look for neat patterns, this is the kind of result
we should expect from investigating dialect boundaries. Moreover, it is unlikely that linguistic
boundaries will neatly coincide with those of administrative units, though in the case of Ayrshire
Ipresented arguments for believing in a strong sense of local identity and loyalty, reinforced by the
strong condensation symbol of Robert Burns as an Ayrshireman.

Dialect

of American English identified by geographical location, such as Appalachian English,

since its use is related primarily to ethnicity not to place. AAVE is a (basilectal) variety
of African American English linked primarily to age and socio-economic status.
4. Some examples of dialect studies3
If Trudgill and Chambers are correct in claiming that dialect is a matter of grammar
then there have been few studies of dialects. There is a limited amount of information
on non-standard grammar (Edwards & Weltens 1985; Trudgill & Chambers 1991),
but most studies of individual dialects concentrate on phonology and vocabulary. The
traditional method of elicitation through questionnaires militated against the collection of much information on syntactic patterns, other than forms that were known in
advance to be variable. One of the most extensive accounts of grammatical use in a
single community is Feagin 1979s study of the town of Anniston (population 31,500)
in Alabama. She was able to find age, gender, and social class differences in features
of the verb phrase, but she was also able to show that all members of the community
basically shared the same system.
In my own work on the dialect of Ayr (Macaulay 1991) I did not set out to describe
the dialect and I would have collected the data differently if that had been my intention. Nevertheless, in the end, I presented a considerable amount of information on
that dialect. I examined phonological, morphological, syntactic, lexical, and discourse
features. I was able to show differences not only in the use of syntactic structures but
also in such discourse features as the expression of intensity. More work paying attention to this range of features needs to be done, including more attention to prosodic
features and voice quality.
Feagin and I both claimed that we were giving an account of a single dialect despite
the variation we found. In towns the size of Ayr (c.50,000) and Anniston (31,500) there
would be few middle or upper class speakers who had not heard most of the forms used
by the lower class speakers or who could not mimic them to some extent (Macaulay
1991:183186). Even in a city the size of Glasgow (c.900,000) middle-class speakers can
enjoy jokes involving lower-class forms of speech (Macaulay 1997a:6169). Descriptions of dialects need to take into account passive knowledge as well. The dialects
of Anniston, Ayr, or Glasgow are not restricted to the kinds of rare survivals that
. This section is not an attempt to review all the work that has been done in the field recently.
That would be a vast undertaking and require much more space than I have available. Instead,
Iwill refer to some examples that I think contribute to an understanding of the notion dialect. The
inclusion of any study does not imply that I think it is the most important, nor does the omission
of reference to a project have any negative implications about its quality or value.

67

68

Ronald Macaulay

Martinet accused dialectologists of being preoccupied with. The dialect consists of

the totality of forms potentially available to all inhabitants of the area. Consequently,
studies of social class, gender, age and other differences complement dialect studies
but do not replace them.
Preston (1989, 1993) in studies of folk linguistics or folk dialectology has investigated non-expert perceptions of dialect differences in the United States.4 He has
shown how respondents are willing to rate speech of their own and other regions on
scales of correctness, pleasantness, and difference and thus provide a mapping of
variation in attitudes toward other forms of speech. Not surprisingly, the area that was
the most salient for respondents from Hawaii, Indiana, and Michigan was the South.
Preston also asked respondents in Michigan and Indiana to identify taped samples
from a series of speakers ranging from Michigan in the north to Alabama in the south.
The respondents were able to do this with some success but seemed to hear more differences closer to home. Prestons folk dialectology provides an alternative approach
to the identification of dialect boundaries which could help complement the results
from surveys of language use but it would not replace them.
5. Some recent developments in the U.S.
Four major recent contributions to dialectology in the U.S. are The Linguistic Atlas
of the Gulf States (LAGS), The Linguistic Atlas of the Middle and South Atlantic States
(LAMSAS), The Dictionary of American Regional English (DARE), and The Phonological
Atlas of North America (TELSUR). LAGS is based on 1121 interviews, conducted in the
years 19681983, yielding 5300 hours of taped material containing approximately 100
million words. The area surveyed ranges east from Texas in the west and south from
Arkansas and Tennessee in the north to Florida. The materials are accessible in several
forms. Copies of the original tapes are in the archives of the University of Georgia.
Two million phonetic strings are available on microfiche as the basic materials from
the tapes and the same items are similarly available on microfiche as a concordance in
orthographic form. The materials will also be made available in a variety of formats on
disk in ASCII files. These will allow researchers easy access to the data base. Finally,
the LAGS materials have been published in seven volumes. The Handbook (1986)
summarizes the information gathered through field work. The General Index (1988)
. It is unlikely, however, that folk linguistics in most countries will display the awareness of
other dialects that is found in Norway. Jahr (1997) refers to a TV program in Norway in which
teams competed to detect on the basis of a persons dialect features, where he or she was from in
Norway (p.367). The contestants had to place the person in the right town, village, or parish. In all
130 rural and urban dialects were presented.

Dialect

summarizes the contents of the concordance. The Technical Index (1989) lists the
contents of the files on disk. The Regional Matrix (1990) outlines the regional dialects
with 7,422 maps. The Regional Pattern (1991) provides mapping a dialect features. The
Social Matrix (1991) outlines social dialects with more than 5,000 maps. The Social
Pattern (1992) identifies configurations suggested by the maps of the Social Matrix.
Itis a remarkable achievement and one whose impact has yet to be fully felt. The compilers of LAGS have produced a magnificent corpus in a variety of forms that should
make it readily accessible for numerous projects. It is now up to those with questions
about dialects to make use of this resource in order to seek out some answers.
The materials for The Linguistic Atlas of the Middle and South Atlantic States
(LAMSAS) were collected sixty years ago but have never been published. They are now
stored at the University of Georgia, where William Kretzschmar is directing a project
to have the records fully edited and stored in database files that will make another
corpus available for future research (Kretzschmar, McDavid, Lerud & Johnson 1993).
As Kretzschmar & Schneider observe traditional dialectology has been slow to institutionalize the benefits of good counting but we now have the computer means, and
we can now explore the ways of serious quantitative analysis (1996:1). One example
of this using the the LAMSAS materials is Johnsons study of lexical change in the
southeastern U.S. (Johnson 1996).
Another major undertaking in the U.S. is the Dictionary of American Regional
English (Cassidy 1985; Cassidy & Hall 1991, 1996). Although the initial notion of this
work was first voiced in the American Dialect Society over a hundred years ago, the
project did not officially begin until 1965. Fortunately, the delay meant that by the time
work started it was obvious that the material would have to be stored on computers,
and that has made the processing of responses quicker and more effective. It has also
allowed the automatic generation of display maps, which are not particularly pleasing
to look at but are very successful in presenting the distribution of variants. DARE is
based primarily on the responses of 2777 respondents who were asked 1847 questions.
Since two-thirds of the respondents were over 60 and only 10% younger than 40, their
responses mainly reflect an earlier stage of the language. As Cassidy observes: Our
field collecting was done just in time before the enormously powerful penetration of
television to every corner of the nation began. Thus a great body of data was saved
from oblivion (Cassidy 1993:104). When completed DARE will provide a firm basis
on which more detailed local projects can build.
The fourth project I wish to mention is also still in the process of completion. The
Phonological Atlas of North America is directed by William Labov at the University
of Pennsylvania. By means of a telephone survey (TELSUR) the project has collected
information on the vowel systems of 607 speakers in the urbanized areas of the United
States. The survey will later be extended to include Canada as well. On the basis of the
information collected so far on minimal pairs and other elicited forms Labov and his

69

70

Ronald Macaulay

colleagues have identified four major dialect regions in the U.S.(identified solely on
the basis of pronunciation): the Inland North, the South, the West, and the Midland.
Three of these areas are characterized by major sound changes in progress: the Inland
North by the Northern Cities Shift, the South by the Southern Shift, and the West by
the Low Back Merger. These patterns are described in detail in Labov (1991) and the
results of TELSUR have supported the original analysis. The Midland is a residual area
with much greater diversity, where individual cities have developed dialect patterns of
their own. There is also a division between the South Midland and the North Midland
with the latter having no single defining feature. However, according to Labov, the
North Midland is not the most unmarked dialect; in his view, Northern Iowa comes
closest to the original pattern of American English. Labov claims that his results show
clear and distinct dialect boundaries, delineating areas with a high degree of internal
homogeneity (Labov, Ash & Boberg 1997:2). He points out that the border separating
the North from the North Midland coincides with the boundary of lexical features in
the Dictionary of American Regional English (Labov et al. 1997:7). He expects later
volumes of DARE to provide further confirmation of the dialect boundaries. Thus we
have a situation in which the work of dialectologists and sociolinguists can be seen to be
complementary in a constructive way. But Labov is not content with identifying major
dialect boundaries. He is also interested in the features that identify smaller areas,
including major cities and other urbanized centers. When completed, the Phonological
Atlas of North America will provide a map of the whole of the United States and Canada
that will enable local investigations to be conducted in the context of the larger picture.
Labov has made dissemination of the results of TELSUR available much more quickly
and easily than is normally the case by posting them on a web site. For a small number
of speakers it is even possible to hear examples of key vowels. In every way it is a magnificent achievement and a major contribution to the study of dialect.
It is obvious from the above projects that recent improvements in technology have
revolutionized dialect studies. Further work in corpus linguistics will lead to new forms
of analysis. The future of dialectology looks a lot brighter than it did fifty years ago.

References
Bailey, C.J. (1973). The patterning of language variation. In R.W. Bailey & J.L. Robinson (eds.) Varieties
of present-day English: 156186. Macmillan.
Bickerton, D. (1973). The nature of a Creole continuum. Language 49: 640669.
Bloomfield, L. (1933). Language. Holt.
Cadiot, P. (1987). Roofless dialects (roofless speech). In U. Ammon, N. Dittmar & K. Mattheier (eds.)
Sociolinguistics: 755760. de Gruyter.
Cameron, D. (1995). Verbal Hygiene. Routledge.

Dialect

Cassidy, F.G. (1993). Area lexicon: The making of DARE. In D.R. Preston (ed.) American Dialect
Research: 93105. John Benjamins.
Cassidy, F.G. (ed.) (1985). Dictionary of American Regional English, AC. Belknap Press.
Cassidy, F.G. & J. Hall (eds.) (1991). Dictionary of American Regional English, DH. Belknap Press.
(eds.) (1996). Dictionary of American Regional English, IO. Belknap Press.
Chambers, J.K. & P. Trudgill (1980). Dialectology. Cambridge University Press.
Edwards, V. & B. Weltens (1985). Research on non-standard dialects of British English: Progress and
prospects. In W. Viereck (ed.) Focus on: England and Wales: 97139. John Benjamins.
Feagin, C. (1979). Variation and change in Alabama English: A sociolinguistic study of the White
community. Georgetown University Press.
Francis, W.N. (1983). Dialectology. Longman.
Gauchat, L. (1903). Gibt es Mundartgrenzen? Archiv fur das Studium der neieren Sprache 111:365403.
Haugen, E. (1966). Dialect, language, nation. American Anthropologist 68: 922935.
Hudson, R. (1996). Sociolinguistics. Cambridge University Press.
Jahr, E.H. (1997). On the use of dialects in Norway. In H. Ramisch & K. Wynne (eds.) Language in
time and space: 363369. Franz Steiner Verlag.
Johnson, E. (1996). Lexical change and variation in the Southeastern United States, 19301990.
University of Alabama Press.
Kloss, H. (1978). Die Entwicklung neuer germanischer Kultursprachen seit 1800. Schwann.
Kolb, E. (1966). Phonological atlas of the Northern Region. Francke.
Kolb, E., B. Glauser, W. Elmer & R. Stamm (1979). Atlas of English Sounds. Francke.
A Kretzschmar, W., V. McDavid, T. Lerud & E. Johnson (eds.) (1993). Handbook of the Linguistic
Atlas of the Middle and South Atlantic States. University of Chicago Press.
Kretzschmar, W.A. & E.W. Schneider (1996). Introduction to quantitative analysis of linguistic survey
data: An atlas by the numbers. Sage.
Kurath, H. (1972). Studies in area linguistics. Indiana University Press.
Labov, W. (1972). Language in the inner city. University of Pennsylvania Press.
(1981). Field methods of the project on linguistic change and variation. Sociolinguistic Working
Paper, No. 81. Southwest Educational Development Laboratory.
(1991). The three dialects of English. In P. Eckert (ed.) New ways of analyzing sound change:
144. Academic Press.
Labov, W., S. Ash & C. Boberg (1997). A national map of the regional dialects of American English.
Macaulay, R.K.S. (1985). Linguistic maps: Visual aid or abstract art? In J.M. Kirk, S.F. Sanderson &
J.D.A. Widdowson (eds.) Studies in linguistic geography: 172186. Croom Helm.
(1991). Locating dialect in discourse: The language of honest men and bonnie lasses in Ayr.
Oxford University Press.
(1997a). Standards and variation in urban speech: Examples from Lowland Scots. John
Benjamins.
(1997b). Ayrshire as a linguistic area. In E.W. Schneider (ed.) Englishes around the world:
159171. John Benjamins.
Martinet, A. (1954). Dialect. Romance Philology 8: 111.
Mather, J.Y. & H.-H. Speitel (19751977). The Linguistic Atlas of Scotland: Scots section. Croom Helm.
McDavid, R.I. (1979). Dialects in culture: Essays in general dialectology. University of Alabama Press.
Milroy, L. (1980). Language and social networks. Blackwell.
Orton, H. & N. Wright (1975). A word geography of England. Seminar Press.
Orton, H., S. Anderson & J. Widdowson (eds.) (1978). The Linguistic Atlas of England. Croom Helm.

71

72

Ronald Macaulay
Pederson, L. et al. (eds.) (19861992). Linguistic Atlas of the Gulf States. University of Georgia Press.
Preston, D.R. (1989). Perceptual dialectology. Foris.
(1993). Folk dialectology. In D.R. Preston (ed.) American dialect research: 333377. John
Benjamins.
Stewart, W.A. (1968). A sociolinguistic typology for describing national multilingualism. In J.A.Fishman
(ed.) Readings in the sociology of language: 531545. Mouton.
Trudgill, P. (1986). Dialects in contact. Blackwell.
Trudgill, P. & J.K. Chambers (eds.) (1991). Dialects of English: Studies in grammatical variation. Longman.
Wolfram, W. & A. Cavendar (1992). Dialect and special-interest domains: Conceptual and methodo
logical issues in collecting a medical lexicon. American Speech 67: 406420.

Dialectology
Georges De Schutter

1. Definition
Dialectology is often defined plainly as the study of one or more synchronic nonstandard language systems (regional dialects, sometimes also regiolects) which form
part of a more encompassing set of varieties, usually called a diasystem. The term
diasystem stands for the complete set of varieties (diachronic as well as diatopicsynchronic) supposed to derive from one ancestor; often a restriction is added in the
sense that such a diasystem also presupposes a common standard language covering
all synchronic varieties.
This type of definition of dialectology appears to be at once too broad and too
restricted in a number of respects. On the one hand, the term dialectology is often
restricted to the comparative study of dialects (including standard varieties), either
synchronical (dialect geography) or diachronical (language change). The study of
phenomena within one single synchronic dialect, be it standard or non-standard,
does not in fact imply any special methodology, and thus belongs to general linguistics.
On the other hand, dialectology often takes into account dialects which undisputably
belong to different diasystems. A well-known example of this is provided by Balkan
linguistics. There has been a long tradition in this area, to study phenomena across
dialects which belong to various Slavic diasystems, to Rumanian, Greek, Albanian
and Turkish. The attempt at compiling a linguistic atlas of the dialects of Europe (Atlas
Linguarum Europeae) aimed at a similar universalist objective.
One further problem with delimiting the field of dialectology is the definition of
dialects as language systems in their own right. As such the notion stands apart both
from that of accents and that of sociolects. The latter terms are applied to language
varieties discerned on the basis of rather superficial characteristics, such as surface
realizations of phonemes or the actual choice between lexical synonyms or morphological and syntactic alternatives.
2. An outline of history
Extensive interest in regional dialects dates from the Romantic period in Western
Europe, but at first this interest led almost exclusively to the compilation of sample texts.

74

Georges De Schutter

The most famous example of this type of work was that initiated by Coquebert de
Montbret during the first French Empire: dialect speakers were asked to write down
the biblical parable of the Prodigal Son in their local dialects. Similar endeavors soon
followed in Germany and the Netherlands.
Dialectology as an independent branch of the sciences of language took its origin
in the wake of two quite different scientific traditions and has been since then constantly nourished by new scientific paradigms.
The first important source was linguistics, first of a neogrammarian, later of a
predominantly structuralist (and generative) signature. Structuralist methods especially have proven fruitful for the development of dialectology, which, in turn, yielded
clearer insights into the make-up and the functioning of natural language systems. As
comparative dialect research usually takes into account closely related varieties, the
impact of even very small changes on the language system as a whole stands out very
clearly. The bulk of work in this tradition covers topics in phonology, morphology and,
to a lesser degree, syntax.
Another field of interest which has given an impetus to the development of dialectology is anthropology: the relation between (the development of) culture and language was a main interest for a number of early dialectologists. It manifested itself
in the first place in the study of the lexicon. In fact, quite a few early dialectologists
believed that comparison of the lexicon of various related dialects would uncover parts
of the evolution of concepts of both material and spiritual culture.
It must be noted, however, that interest in the evolution of the lexicon of dialects
is not exclusively related to this anthropological background. The study of the geographical and/or social distribution of word types may also be a by-product of the
structuralist mainstream of dialectology.
One objective may be to get a grip on the existence of structural exceptions, e.g.by
providing evidence that exceptions in one system are loans from other geographical or
social dialects, in which they are the result of regular application of language laws. The
explanation is thus looked for in the phenomenon of the migration of words across
varieties of the diasystem; a good example of this is the word beuk (beech) in most
western Dutch dialects, including the standard variety: the phonological form points
to northeastern dialects of the Dutch diasystem, and one of the hypotheses in connection with this word form is that it came to the west together with the tree itself.
Another domain is the study of lexical gaps, the fact that words vanish from the
language. A well-known example is Gillirons gallus-cattus case (cf. Pathologie et
thrapeutique verbales, 19151921), which led to the notion of homonymic conflict:
if different lexical items (threaten to) fall together by the application of phonetic laws,
one or even both of them tend(s) to be replaced by (a) heteronym(s).
A third aim is to offer insight into the internal structure of lexical fields (German
Wortfelder). The focus, then, is on dialects which display differences in the distribution

Dialectology

patterns of common words over concepts in a semantic field. If the dialects are regionally
adjacent, the result of conflicts between lexical systems may be studied as well (often
inhabitants of border regions between dialect areas avoid words which have different
meanings in the core dialects).
The prevailing interest in the interaction of competing language systems and the
awareness that regional dialects are often, in a naive way, connected with social backwardness, or, as romanticists used to put it, naturalness, finally led scholars to pay
attention to communicational aspects of dialect use. This gave rise to communicational
dialectology, a branch which makes use of sociological and sociopragmatic methods
and ends up in work that is very much akin to other types of sociolinguistic research.
Interest naturally shifted from dialects as monolithical systems governed by strict
structural laws, to dialects as complex tools with a wide spectrum of lexical and
grammatical variants determined by social and/or situational parameters. In this way
the focus within dialectology shifted more and more to the social gradation of competing forms within systems in evolution. At the same time the relations between
social varieties of the same diasystem were foregrounded: the factural and emotive
relations between various types of regional dialects (e.g. urban vs. rural), regiolects
(i.e. varieties other than the standard one(s), functioning in a more or less extended
part of the language community), sociolects and standard varieties were gradually
being taken into account, along with the attitudes of both speakers and non-speakers
towards those varieties. Especially objective appreciation (suitability in more or less
formal situations, pleasing or repelling characteristics, etc.) became an object of study.
In fact, on account of this development, the dividing lines between dialectology and
sociolinguistics were and are getting blurred.
From the 1960s on, interest thus gradually shifted from dialect system to dialect
use. This resulted, among other things, in new collections of electronically recorded
dialect data, often more or less free conversations or monologues. The aim was, beside
(once again) preserving samples of dialects on the verge of extinction, to offer research
material for the study of non-standard speech. Here, again, we sometimes find the old
idea of the superiority of non-standard dialects, this time as a preferential source of
insight into natural ways of communication, reputedly less disturbed by normative
prescription than those adopted by speakers of standard varieties.

3. Aims of the investigation

As a consequence of this multiple origin and the constant adaptation of aims and
methods to the ever-changing scene of linguistic, social and anthropological sciences,
both modern dialectology and its predecessors display a wide variety of research topics.

75

76

Georges De Schutter

Though many (perhaps even most) studies combine linguistic, historical and sociological aspects, the following objectives are found in fairly pure form.
3.1 The impact of history on the origin and evolution of languages
One of the theses of dialectologists is that the spread of changes throughout a language
area, and even through different language areas belonging to one cultural or political entity, is a by-product of the communitys evolution, i.e. of the external circumstances in which the language developed. The spread of innovations is thought to be
promoted by the radiation of strong cultural, religious or economic centers (political or
ecclesiastical seats, market places, etc.), whereas it is slowed down, or even stopped,
by political, or religious boundaries and by most types of natural geographic boundaries
(mountain chains, forests, large water surfaces, but usually not rivers). The history
of language communities is reflected both in structural (phonological, morphological,
syntactic) and lexical phenomena. Lexical or structural oppositions are often geographically divided by clear-cut boundaries: isolects, isoglosses, isomorphs, isotagms
(bearing on lexical, phonological, morphological and syntactic oppositions, respectively). Though complete overlap of intra- and extra-linguistic boundaries is rare, they
often converge considerably. Apart from being products of history, dialects were also
considered a mirror of historical evolution. As was hinted at before, romantic cultural
movements have always taken a vivid interest in the ways and customs of rural communities. This also, or even in the first place, applied to language varieties. Rural dialects
were thought to be free from corruption by extra-linguistics facts such as language
planning, infliction of normative rules by authorities, etc., and were thus supposed
to reflect the authentic historical developments of the communities in which they
were spoken, more clearly and more directly than urban varieties, including standard
languages. Dialectologists such as Jaberg (Italy and Switzerland), Van Ginneken (the
Netherlands), and many others saw rural dialects as shrines containing the residue of
natural civilization. This must not be interpreted in too narrow a folkloristic sense.
At least the leading researchers in this tradition tried to gain insight into the history
not only of material civilization, but also of spiritual and religious traditions: where
and when they originated, how they spread, and which part they played in the development of local culture.
3.2 Linguistic reconstruction
Dialects played an important role in historical linguistics, in at least two ways. First
of all, there was the idea that the original structures and the relations between these
are reflected more faithfully in natural languages than in standard varieties, the latter
being more open to unsystematic incorporation of foreign elements.

Dialectology

Second, the existence of different language systems, deriving in a natural way from
the same underlying set of rules, is a further advantage for the study of language change.
In this respect structuralists formulated the hypothesis (for many of them even a kind
of dogma) that synchronic adjacency (German das Nebeneinander) is often a mere
reflection of diachronic succession (das Nacheinander): the geographic distribution of
phenomena may be an indication of the way linguistic development took place; innovations mostly start in one dialect (area) and gradually spread across other dialects
(cf. the second Germanic consonant shift). As one proceeds from what is called the
epicenter to the periphery of application of a systemic change, either fewer structural
elements are struck by the phenomenon, or a certain number of lexical elements are
affected, others remaining untouched (the phenomenon of lexical diffusion).
3.3 The study of universals of language, especially implicational scales
The study of closely related dialects can be very useful as a complement to research in
totally unrelated languages. All deviances occurring in related dialects can be mapped
against the broad set of similarities in the dialects at stake. In this way the interrelation
and relative (in)dependence of phenomena may be illustrated more convincingly and
insightfully than in traditional universalist work, which is based on the comparison
of preferably maximally different unrelated languages.
This also applies to pragmatic universals, of course. The comparative study of
aspects such as formality, politeness in language, etc. is very well possible in different
varieties of one diasystem. As a matter of fact, dialects do not only differ from other
dialects (whether standard or not) in the referential tools at hand (grammar, lexical items such as nouns and verbs), but also in a number of non-referential aspects,
including attitudinal particles, modal auxiliaries, politeness formulae, and other means
of guiding social communication on a pragmatic level. For the study of these elements,
especially of particles, differences in closely related languages may shed a clear light on
the impact of oppositions in very specific fields, and on the functioning of the system
as a whole.
3.4 Sociopragmatic and attitudinal aspects
Sociopragmatic and attitudinal aspects did not receive much attention until the last
few decades. At least three aspects are now clearly in focus. First, there is significant
interest in the (system-external) study of the overall functionality of geographical
varieties in a language community, and of attitudes in different groups within this
community. For this aspect, of course, the methodology applied to geographic, ethnic
or social dialects, also applies to the distribution and the appreciation of social accents

77

78

Georges De Schutter

on the one hand, of competing languages on the other. It is simply part of the sociology
of language.
Second, there is the (system-internal) study of interferences (grammatical and
lexical) between varieties, adaptations of one variety to another, especially in diglossic
situations, e.g. in urban areas. Here, again, there is no apparent difference between
this type of research and the study of different languages (diasystems) in the same
types of situations.
Third, the sociopragmatic study of dialects also comprises quantificational aspects.
Sample data of geographical, ethnic or social varieties of a language may be examined
in order to establish the degree of dialecticity, i.e. the rate of divergence of a dialect
vis--vis an objective reference variety such as the/a standard variety. Most attempts
in this field have proven unsatisfactory: the researcher is compelled to stick to stereotypes both in the classification of data (as belonging to a specific text type, e.g. interview, informal discussion, etc.) and in the characterization of participants (e.g. on
a hierarchical basis, between high and low). Extrapolation to the dialect as such
proves practically impossible.
3.5 Communicative aspects
One of the main problems conversational dialectology has to deal with is that of the
social values associated with (elements of) competing systems. One view in this respect
is that there are more or less fixed correlations between linguistic facts and personality
factors determining the speakers social life (sex, age, profession, cultural background,
etc.). Another one (not necessarily in contradiction with the first) is that correlations
also exist between the sets of linguistic tools at hand for one speaker or for a well
defined group of speakers, and the situations in which language is used, e.g. casual
speech, more or less formal monologue, reading aloud, etc. The decisive impetus to
this type of research was given by Gumperz (e.g. 1982). The following orientations, at
least, must be discerned.
First, there is research focusing on the speaker. It aims at assessing criteria such as
knowledge and awareness of norms. It is usually held that divergence from the norms
valid in a variety is caused by lack of knowledge. This belief has been severely shaken
by recent research: mixing up elements from different sets need not be accidental at all,
it may very well be used intentionally, both as a language game and as a very personal
way of adding expressivity to the message.
Second, there is research dealing with the hearer. In this case the main focus
will be on criteria such as intelligibility and acceptability of the language and, more
important, on the elements causing different attitudes towards it. One very interesting
topic in this respect is the relation between language and ethnicity, as manifested in
the (positive or negative) appreciation of often closely related dialects across state

Dialectology

boundaries (many new nations in Africa, the European countries where German is the
mother tongue, Latin America).
Third, there is also research on the process of communication itself, such as the
matching of stimuli and responses, evasion strategies, etc. In this type of research
the conversation itself, or rather the relation between the participants, is taken as the
principal starting point; the researcher tries to look behind the actual motives of the
interactants in shifting from one (set of characteristics of a) variety to another one. At
the same time, the interactants relation is taken to be liable to shift, and this is considered to be a worthy topic of research as well. One problem for European traditional
dialectology, with its strong positivist (empiricist) background, is the impossibility of
generalizing from casual data. This is why, up to this moment, this type of conversation
analysis is almost absent from the dialectological scene, at least in Europe.
4. Dialect atlases
The comparative character of dialectology is best reflected by the existence of large
linguistic atlases. From the very beginning of the scientific study of dialects, large collections of linguistic material were set up and the evidence was brought together in dialect
maps. In fact, much theoretical work in dialectology was concerned almost exclusively
with the requirements of the geographical visualization of data from various dialects.
There are, fundamentally, two major traditions in this respect, initiated by two of the
earliest great endeavors which may, even now, be called their prototypes.
On the one hand, the Atlas linguistique de France strives to show the bare facts,
leaving any interpretation to the objective observer: the data are simply written on the
geographical coordinates of the community using them. The Sprachatlas des deutschen
Reichs, on the other hand, offers maps consisting of interpretations of the data by the
researchers. Here all data are represented by a set of abstract symbols or are simply
lumped together, forming a lexical, phonological, etc., area covered by the same item.
Of course, the symbols forms, colors, etc., may suggest both relationships and contrasts between data, not all of which will be beyond discussion. The conciseness of the
symbols allows the dialectologist to bring together many more data items in a single
map. That may be the reason why this system was adopted even for the fairly recent
Atlas linguarum Europeae.

References
Busch, W., U. Knoop, W. Putschke & H.E. Wiegand (eds.) (1982). Dialektologie. De Gruyter.
Chambers, J.K. & P. Trudgill (1980). Dialectology. Cambridge University Press.
Gumperz, J.J. (1982). Discourse strategies. Cambridge University Press.

79

80

Georges De Schutter
Jaberg, K. & J. Jud (1928). Der Sprachatlas als Forschungsinstrument.
Pop, S. (1950). La dialectologie. Recueil Universit de Louvain.
Schouten, M.E.H. & P.T. van Reenen (eds.) (1989). New methods in dialectology. Foris.
Shuy, R., W. Wolfram & W. Riley (1968). Field techniques in an urban language study. Center for
Applied Linguistics.
Trudgill, P. (1983). On dialect. Blackwell.

Evolutionary pragmatics
Wolfgang Wildgen

1. Pragmaticism, pragmatics, adaptation and the evolution of language

Historically there is a link between Darwins theory of evolution (1859, applied to
humans 1870) and discussions in the Metaphysical Club at Harvard, which began
meeting in 1872 and had William James and Charles Sanders Peirce as members.
Their point of view was that pragmatics is Darwins theory of natural selection applied
to philosophy. In a speech delivered in Harvard in 1872 Ch. S. Peirce sketched his
Pragmatism as a philosophy based on the practical consequences of intellectual
operations. The term pragmatic refers to Kants Anthropologie in pragmatischer
Absicht. From its beginning, pragmatics had therefore a strong link to anthropology
(cf. Kant) and evolutionary theory (cf. Darwin) with its central concept of adaptation
(cf. Verschueren & Brisard 2002). stman (1988) points to different levels of adaptation.
One has to distinguish adaptive processes found in animals which shape instincts:
Instinct can be seen as a response to the pressures of natural selection, and it thus
functions as a means of preservation of the species (stman 1988:11), and adaptation can have different goals: adaptation to a changing ecology, adaptation to changes
in the social organization and cultural context of populations and finally adaptation to
new standards of language use. When Whorf describes the world-view of the Hopi, he
can point to a divergence in human cultural adaptation, which results from a separation of the Amerindian population from the Europeans some 40 to 50,000 years ago.
In sociolinguistics the separation of social classes creates different codes as modes of
communication in different social networks (cf. the work of Bernstein and others). On
a much shorter time scale, every conversation creates meanings in addition and/or
opposition to speakers intentions (stman 1988:10). In the following sections, we
will focus on the evolution of language and its pragmatic dimensions before the last
out-of-Africa migration and the separation of the linguistic families.
While for the pragmatist a gradient transition between communication in animals
and language in humans is evident (as it was for Darwin), some currents in theoretical linguistics take the opposite position. Fitch, Hauser and Chomsky (2005) focus
the question of language evolution on an abstract core of computational operations,
central to language and probably unique to humans (Fitch, Hauser & Chomsky
2005: 180), and they try to abstract the myriad component mechanisms (p. 181)
which are not unique for humans. What is left is called FLN: faculty of language in
the narrow sense. As they say themselves, FLN could possibly be empty if empirical

82

Wolfgang Wildgen

findings showed that none of the mechanisms involved are uniquely human or unique
to language. For Fitch, Hauser and Chomsky (2005), the best (and currently only) candidate for a component of FLN are computational capacities of recursion (p.204).
As only synchronic adaptation is accepted as empirical evidence, all explanations
arguing for an adaptive origin of this core feature are dismissed and thus any evolutionary explanation of human language is considered to be unscientific (not testable).
Jackendoff and Pinker (2005) argue that recursion is also found in the human visual
system and thus would not be specific to language (i.e. not part of FLN, which as a
consequence would be empty). Moreover the narrow syntax-view of language in
Chomskys tradition is rejected as scientifically unproductive.
In the following we join Darwins basic intuition that there is a gradual (possibly
punctuated) transition between animal communication/cognition and the linguistic
competence of humans, i.e. our object is the faculty of language in a broad sense.
Nevertheless, specific human abilities demonstrated by the use of language remain the
explanatory goal of the endeavor.
The evolution of human cultures and languages seems to be due to rather quick
developments (relative to the phylogenetic scale). Two major periods of rapid emergence are plausible:
The first is linked to the emergence of lithic technology (Homo Habilis) and to
the proto-species which was later distributed in Africa, Asia and Europe: Homo
Erectus (and its African variant Homo Ergaster). One can postulate that the core
of this evolution took place around 2 my BP. (2.4 to 1.6 my).
The second concerns the archaic Homo Sapiens prior to the second Out-of-Africa
migration. We may fix the core of its evolution to the period 200 ky BP (currently
the oldest bones associated with Homo Sapiens are 150 ky old, but genetic calculations based on the diversity of current populations lead to a time-span between
400 ky and 150 ky)).1
Comparative behavioral research shows that many cognitive and communicative
skills were pre-existent to both species and thus to the emergence of language. The
list contains at least the tendency towards bigger brains, upright locomotion, ad hoc
tool use, call patterns, manual dexterity, and specific brain functions (asymmetry of the
parietal areas, forebrain extension, mirror cells). Darwins (1969 [1872]) assumption

. New analyses of the genome of three Neanderthal men and a specimen of humans living in
different parts of the world showed that all non-African populations share genes with Homo
Neanderthaliensis. They intermixed probably in the phase when Homo Sapiens left Africa (between
100 and 60 ky). Cf. Green et al. (2010).

Evolutionary pragmatics

that human language arose continuously from animal expressive and communicative
behavior has to be revisited, insofar as different neural pathways are used by primate
calls and human language. Nevertheless human language did not evolve out of completely organized grammars (and their semantics and pragmatics). It seems rather that
the ecology, the actions of (pre)humans in their ecology, the technologies evolved since the
Homo Erectus and the cultural/social advances they allowed, are the proper background on
which human language could take shape. In short, the pragmatics of prehuman communication is the platform on which human language was built. For this reason it seems
reasonable to call the corresponding field of research evolutionary pragmatics.
We shall first expose the evolutionary signification of traditional sign functions,
then treat the role of artifacts and lithic technologies for the evolution of symbolic
behavior and finally discuss the transitions from ecological to cultural pragmatics.

2. Sign-functions and their evolutionary significance

Cassirer as well as his contemporary, the psychologist Bhler (cf. his Sprachtheorie,
1934), assume a critical limit between human and animal sign behavior on the level of
sign-functions. For Cassirer the symbolic organization of thinking (giving symbolic
form to pre-symbolic thought) demarcates the transition between nature and culture.
Bhlers and Cassirers views are part of the common attitude to Darwinism between
1920 and 1935. Neodarwinism, also called Modern New Synthesis, which has a solid
foundation in genetics, was created between 19361947. Studies on the intelligence of
higher apes had been initiated in the twenties by Khler and were further developed in
the decades following World War I. It is therefore necessary to reevaluate the proposals
by Cassirer and Bhler in the light of more recent results. I shall elaborate on Bhlers
proposals in view of an evolutionary structure underlying human sign functions.
2.1 The triad of sign functions
Bhler distinguishes three major functions of signs: representation (Darstellung),
expression (Ausdruck) and appeal (Appell). In his view, the function of representation, i.e. systematically organized reference to the world (including sign users) is only
achieved by human language, whereas the two other functions are already realized
in animal communication. Therefore, semiotic signs (as triadic systems of functions)
are different from non-semiotic signs (= signals). Later Jakobson (1981:2223) further
elaborated this triad and made a distinction between referential function, emotive function, conative function (these correspond to Bhlers original triad), phatic function
(establishing contact), metalingual function (telling us something about the code),

83

84

Wolfgang Wildgen

and poetic function (concerning the form of the message). The three functions which
Jakobson has added are self-referential, insofar as they concern aspects of the basic
triad; we may call them meta-functions (cf. also Wildgen 2009a). In Figure1 the basic
triad of Bhler is illustrated and the metafunctions (Jakobsons three further functions) are added. The function of representation is only clear in some examples of
animal behavior and remains very context dependent.

Meta-functions (phatic, metalingual, poetic)

Representation
Human language
Semiotic sign
Animal communication
Non-semiotic sign
Expression

Appeal

Figure 1. Pragmatic functions and the hierarchy of signs

For Bhler, functions (aims, intentions) are kinds of vital needs and thus presuppose the level of life (of animals). If such needs (or instincts in traditional termino
logy) are generalized beyond animals to humans, a higher level of generalization
can be reached. In fact, the cyclic nature of the schema already became apparent
in Jakobsons additions, insofar as any meta-level generates further meta-levels (ad
infinitum). But even Bhlers original triad is as such the product of a cyclic process
and this cyclic nature makes the triad interesting for a dynamic theory of language use
and language evolution.
2.2 The evolutionary interpretation of the triad of functions
The discussion on primary functions of language is part of a larger discourse on vital
needs of living beings linked to the question of instincts and their classification (cf.the
discussion on pregnancies in Thom 1983:264ff., Wildgen and Plmacher, 2009
and Wildgen, in print 2010). Moreover, the triad implies some kind of intentionality
(i.e.a goal which is perceived and realized) and it is therefore rather anthropomorphic,
which from the start does not invite its application to animals, which are at a low
level of consciousness or social organization. The last two functions, expression and
appeal, are strongly linked, because the use (meaning) of expression asks for some
receiver and appeal is without effect if no content can be expressed. We can use the

Evolutionary pragmatics

label social communication (social calls, grooming, body postures, etc.) as a coverterm for both, and distinguish it from functional referentiality (which first appears
in the alarm-calls of e.g. velvet-monkeys). This simplifies Bhlers triangle to a linear
opposition. If representation is in its first stages already present in socially organized
hominids (or even in monkeys), the transition to humans concerns mainly:
the enrichment of representation, i.e. the lexicon and via self-organization the
syntax; and
the emergence of meta-functions.
The most prominent case of meta-function concerns propositional attitudes and
explicit performatives, cf.:
1. I believe that a snail is in the tree.
2. I tell you that a snail is in the tree.
The content of the propositional attitudes and the preformatives: I believe , I tell you,
does not concern a state of affairs (in/about the world); it is rather about the speaker,
his audience and their social relation. Therefore, it concerns the level of meta-social
communication.
If in a further step one assumes that representation emerges from ecological cognition (categorization of an ecology) and expression/appeal from some structure of
the group (primitive, non-conscious social categorization of behavior), one obtains
three inclusive levels, where the inner circle is reached by all animals with a social
organization and specific reactions (perception/motor control) to their environment,
the middle circle concerns animal communication with a minimal reference to the
context and the outer circle encompasses humans (and possibly some primates with
self-awareness). The functions in Bhlers triad emerge from ecological categories and
from social categories already apparent in animal behavior. They ask for a first level
of self-referentiality and consciousness. The meta-levels (cf. Figure2) go beyond the
emerging self-consciousness and open a range of new behaviors. In this sense the
function-triad implicitly contains a cyclic dynamics, in which basic functions of ecological adaptation and social organization (level 1) emerge (via some hypercyclic process which should be further specified) to the new duality of representation versus
expression/appeal, which is characteristic of higher primates and humans and probably fully developed in a (proto)language with a basic lexicon. The next step was only
reached when self-consciousness and meta-cognition emerged. This presupposes
the evolution of control-functions probably situated in the forebrain of humans. The
language and other symbolic forms latent in Homo Sapiens populations (core time
200ky BP) were further developed via cultural evolution in the last 40,000 years.

85

86

Wolfgang Wildgen

Metalingual function

Poetic function

Representation
Ecological categories
Social categories
Expression/appeal
Phatic function
Figure 2. Three levels for the emergence of sign functions

As the processes schematically considered in the last sections belong to macroevolution, one must ask for the Darwinian criteria of fitness which have driven such an
evolution. It is clear that every new level of emergence has increased the (mental and
social) costs and therefore one must also find a benefit which can at least compensate
for such costs (or which can better make them a profitable choice).
2.3 Selective value of communication and symbolic behavior
All functional models of communication should ask for the survival value of communication, because in a Darwinian framework only the fertility and degree of survival
of the species (not its cleverness) counts. As communication is a type of informationsharing, a concept of (strong) reciprocity is needed. Under what circumstances did
(reciprocal) sharing of information pay off? According to Fehr and Henrich (2003:3),
A person is a strong reciprocator if she is willing (i) to sacrifice resources to bestow
benefits on those who have bestowed benefits (= strong positive reciprocity) and (ii)to
sacrifice resources to punish those who are not bestowing benefits in accordance with
some social norm (strong negative reciprocity). That is, strong reciprocity is produced
by between-group selectivity because groups with disproportionably many strong
reciprocators are better able to survive (Fehr & Henrich 2003:2829). Within-group
selection favors egotistic behavior. They argue that a population which is often in
danger of extinction and must therefore reorganize (and include strangers) will produce a relevant amount of strong reciprocators. As under normal conditions (without
the inclusion of many strangers), egotistic behavior is favored, the species will in the
long run assemble a stable amount of reciprocators and egotists, i.e. the population is
ethically bivalent. The sharing of information on the ecology, on ones own mind and
on social relations (expression/appeal) would follow from strong reciprocity. It also
enforces a level of truthfulness of symbolic behavior. Cheating and lying by means of

Evolutionary pragmatics

symbolic forms is, however, an alternative corresponding to the within-group selection of egotists. The equilibrium of both strategies and its stability is a phenomenon
which asks for further elucidation.
Another type of functional explanation of language origin based on social evolution and the selection on groups rather than on individuals has been proposed by
Dunbar (1997). He found that chimpanzees employ 20% of their time in grooming.
These practices are necessary to uphold social solidarity, social roles (hierarchies),
to control conflicts, etc.; i.e. grooming is a semiotic activity, a ritualized behavior
abstracted from mutual hygiene. In bonobos, sexual activities are also ritualized for
social purposes. Dunbar argues that the percentage of time spent on grooming activities
depends on the size of the group. If the social organization of the group tends to larger
communities, these techniques of solidarity and social peace become energetically too
expensive. Vocal communication, chatting, simply construing vocalized contexts of
solidarity is an alternative. The most proficient actors in social communication get
dominant roles in the tribe and reproduce at a higher rate. A run-away process makes
this competence desirable and creates the necessary social power. Very soon a population may be organized by selection on communicative, i.e. linguistic competence.
In more recent research (Dunbar 2002) evolutionary criteria are applied to study
the mating choices visible in personal advertisement and in history (marriage records
in Eastern Frisonia, conflict in a Viking society, and others). Although cultural rules
govern the transmission of choices over many generations, the choices themselves
follow criteria of an evolutionary game with the number of offspring and their chances
for survival as guiding criteria.
From these discussions it becomes clear that the functions language may fulfill are
not a constant in human evolution; their change and differentiation could drive the
evolution of human language. In Section3 I will rather turn to the evolution of human
symbolic forms (cf. Cassirer 1957 [1923]) and ask whether more practical symbolic
forms, like technologies and art, are able to shed some light on the evolution of our
language capacity (cf. Wildgen 2004: Chapters 46).
3. C
 an the pragmatics of tool production and tool-use tell us something
about the origin of language?
The first stone tools were produced around 2 my; they make up the so-called pebble
culture.2 The pebble culture requires the use of a stone or bone to chock (another)stone,
. Chimpanzees may use a stone to open a nut; cf. Chapter 4 and current research in the group
directed by Christophe Boesch at the MPI Evolutionary Anthropology in Leipzig. Table 1 in
Boesch and Tomasello (1998:593) classifies the semiotic behaviors in six chimpanzee populations. The group-specific learned behaviors are also called a culture.

87

88

Wolfgang Wildgen

in order to produce a sharp edge on the pebble; i.e. the tool is used to produce a specific
shape and is fitted to a large number of uses. Probably other materials (bone, wood,
and fur) were in turn shaped using the primitive stone tools.
If fire had to be conserved (as in populations found in Tasmania and Australia,
which conserved rather than reproduced fire), the process of fire had to be controlled.
In both cases, it was necessary to master control of causation and instrument use (with
an iteration of processes of cause-effect control), as well as its precondition a representation of possible effects, shapes and functions. The Homo Faber, as Bergson
called man at this stage, had the cognitive abilities for symbolic representations. The
question is: Did he use phonetic language to express these representations, or gestures,
or neither of these? Some authors favor a motor origin of language, and thus stand
in the tradition of Condillacs langage daction (cf. Hewes 1977, who distinguishes a
gestural/semantic and a full vocal language; and Quiatt & Reynolds 1993:266ff.). In
this perspective, the (proto)language of Homo Erectus populations would have been
gestural (with holistic phonation as a supplement).
Artifacts are not only hints at the cognitive level of humans, they are also linked
to social life. In order to produce artifacts and to keep fire, a socially organized exploitation of the environment, a division of labor and a mode of social distribution of
products must be in place. This requires rules of collective behavior, and language is
the prototype of rule-governed social behavior; it not only helps to represent and enact
social behavior, but it is the central symbolic representation of social behavior (cf. also
Habermas 1982 theory of communicative action).
3.1 Instrumentality in higher mammals and man
The use of instruments and the goal-oriented adaptation (manufacturing) of tools
can be observed in many orders of animals: ants (insects), birds, and mammals all use
simple instruments. In some cases, this allows them to access difficult areas of their
body (elephants) or to reach under surfaces. Chimpanzees shape twigs to facilitate
fishing for termites in termite-hills (cf. Immelmann 1979:128). The use of instruments
may be inborn and even the evolution of limbs may be connected to instrumental
functions, i.e. limbs are shaped evolutionarily to adapt to specific instrumental functions. Thus, primate and human hands take over functions originally located in the head
(mouth) for attack, defense, preparation of food, for mastication, etc. Our gestural
language, facial expressions, and vocal language presuppose a kind of instrumental
evolution of the human (and hominid) hand and face (cf. Wildgen 1999b for the synergetics of hands and eyes).
The development of tool-use and tool making implies learning, social imitation or
even teaching. Tembrok (1977:186187) distinguishes six levels, or stages:
1. ad-hoc tool-using (but cf. Davidson & Noble 1993)
2. purposeful tool-using

3.
4.
5.
6.

Evolutionary pragmatics

tool-modifying for an immediate purpose

tool-modifying for a future eventuality
ad-hoc tool-making
cultural tool-making

The last stage, cultural tool-making, can only be observed in hominids and in man.
In a certain sense, human cultures are represented by the production of permanent
tools, the techniques of their usage and the social organization enabling and supporting
their use. The precise use of tools becomes apparent in the throwing of shafted handaxes, and later in the use of arrows.
In the evolutionary line of primates, tool-use is reported both for new world apes
and old world apes. The former show only the behavior of throwing objects (from
above down to the bottom of trees) in attack and defense, whereas the latter show a
higher diversity of tool uses (cf. Becker 1993:79110). Rather sophisticated tool-use
with beginning tool modifying is reported by Boesch (1993), who describes the nutcracking behavior of wild chimpanzees of the Ta National Park (Cte dIvoire). The
animals transport both nuts and hammers to roots, which are used as anvil. As stone
hammers are rare and necessary to crack very hard nuts (Panda oleosa), they are transported and preserved. Wooden hammers may be shortened using fallen branches until
they fit. Infants must learn the use of tools, and different ways of passing on the proper
method of use have been observed: stimulation (e.g. leaving the hammer near a nut),
facilitation (providing good hammers and intact nuts), and active teaching (Boesch
1993:173174).3 Another type of tool use by chimpanzees is called leaf sponging,
i.e.wild chimpanzees drink rain water from the hollow of the trees using leaves.
Although not all chimpanzees in all ecological environments show these types of
tool use, one can say that they are able under proper circumstances to develop a system
of stable tool use and even tool modifying. A moderate amount of teaching of tool-use
is possible without the use of language but complicated actions or their perfect enacting
require special linguistic tools; this is clear in the case of normal musical education
or high level athletic training. A simple level of tool-use and tool-making does not
require language and the immediate question is whether language was a necessary
condition for the further evolution of tool-use, beginning with stage four in the list
above, or whether the general (social) evolution, which demanded an enabled level
of cultural tool-making, had the existence of a language as a (social) precondition.
Athird possibility would be that tool-making at stage four demands planning beyond
the present and at further stages the control of a series of goal-oriented activities,
i.e.in a sense a syntax of manual activities. The production of tools becomes a part of
a larger set of social practices, i.e. tools found by archeologists are only indicators of a
. The oldest nut-crackers of chimpanzees found are dated to 4300 y BP; as no human population
lived in the neighbourhood at that time, the technique could not be borrowed from humans.

89

90

Wolfgang Wildgen

very complex social and cognitive interaction. Thus stone tools of a certain material
and size presuppose knowledge about places where one finds the material, a mental
geography of proper resources. The stone tool in use can help to shape other tools of
wood, horn or bone; these again are helpful in manufacturing clothes, parts of the
furniture and dwelling.
On this view a stone tool is only the single remnant of a whole system of cultural traditions, which were learnt by children, taught by adults and assembled in the
memories of the older members of the clan together with the stories of the family and
the clan (of the world and the spirits possibly). One can easily imagine such a social
complex if one considers the embedding of basic manufacturing techniques into the
community life of Australian aborigines (cf. Reynolds 1983).
Another key to the evolution of tool-use and language is possibly cerebral lateralization, which is a long-range tendency in primate evolution:
For example, hemispheric specializations similar to those that characterize Homo
Sapiens appear to be present in macaque monkeys (Macaca) who are left-hemisphere
dominant for processing species-specific vocalizations [] and right-hemisphere
dominant for discriminating faces.
(Reynolds 1983:224)

In the course of the evolution towards man, the left hemisphere subsequently became
specialized for right-hand manipulation and bimanual coordination. Thus the evolution
of manual skills was responsible for the cognitive ability of planning and coordinating
the motion pattern of hands. In parallel, the anatomy of the hand changed and as
archeologists have discovered enough bones of hands, one can deduce from the characteristics of these bones, that
the Australopithecus Afarensis already had a higher mobility of the hand in comparison to chimpanzees living in that period, but that there remains a clear qualitative difference compared to modern humans and,
the Homo Neanderthalensis of Ferrassie 1 and 2 has specific features which do not
coincide with those found in humans, but the mobility of their hands was presumably at the same level (cf. Piveteau 1991:62 ff.).
The parallel question for an archeologist is: Did Australopithecus Afarensis or Homo
Erectus make tools (beyond level 3, mastered by chimpanzees; see above) and was
Homo Neanderthalensis as fit for tool making as the Cro-Magnon man was?
The earliest tools are dated to about 2 my BP. They were found in the Olduvai
Gorge (East Africa) and show a variety of forms of flaking using pebbles, which had
been brought from other places to the sedimentary context in which they were discovered. The basic technique of stone flaking had been discovered and elaborated to a
culture. For these cultures, the corresponding findings of human bones received the
name Homo Habilis.

Evolutionary pragmatics

The next stage is called the Acheulean industry and is related to the Homo Erectus.
The shape of the bifacial hand axes is (at least locally) standardized (cf. Davidson&
Noble 1993:370371). The archeologists are still debating whether the hand axes or
the flakes (or both) were the tools intentionally produced. The stone-industries of late
Homo Neanderthalensis (Mousterian industry) improved (perhaps out of rivalry with
the Cro-Magnon man) and reached a similar level.
3.2 Is tool-making a pragmatic source of propositional semantics?
Continuing in the line of coevolution of visuospatial scenarios and cognitive-semantic
competence, we can compare tool-making scenarios with schemata for simple sentences. The underlying hypothesis is that the semantic structure of sentences is prefigured in visuospatial scenarios as those mastered by early toolmakers (Homo Habilis,
Homo Erectus). Such a hypothesis is also corroborated by comparative research on
the cognitive development of humans, chimpanzees and monkeys. By testing physical
cognition (causality) and logico-mathematical cognition (classification) it was shown
that not only are humans quicker in their development reaching the second (higher)
cognitive level, but also that humans develop both types of intelligence in parallel,
whereas in chimpanzees physical cognition comes first and overlaps shortly with logical
cognition around the age of two. In humans the physical intelligence, which is dominant in tool use, can thus co-evolve with logical intelligence. This could be a major
precondition for the acquisition of language. At the same time this enables a higher
level of tool-use, including several steps, intermediate goals and subtractive (negative)
techniques in tool manufacturing. The basic script of tool manufacturing contains the
following schemata.
1. Seeking for materials (this may include the cultural transmission of knowledge
where the materials may be found, and even trading of materials).
2. Using both hands, such that one hand fixes the material, which has to be shaped,
and the other controls a tool used for shaping. This means the holding of both
objects and the control of a stroke of the bone-tool on the stone.
3. The products of tool making in the late period (about 3010 ky BP) were highly
differentiated and served many purposes.
4. The tool is adapted to specific contexts; it becomes the blade of a knife, the point
of an arrow, the body of an ax, etc., or it is used to perform one phase of a process,
e.g. cleaning the fur of an animal; the fur is already the result of a longer goaloriented process beginning with the hunting of the animal. If a social distribution
of functions exists, the tool-producer may exchange his product for food or other
tools. It becomes an object of value. The mastering of tool-production allows the
production of cultural objects and art; these may again become objects of value.

91

92

Wolfgang Wildgen

Elaborated tools and objects of art show geometrical abstraction (triangles, symmetrical or asymmetrical shapes) and iconicity (with abstraction).
5. A further stage produces pictures (signs) of the hand, the instrument which
shapes tools.
The last stage points to a first cycle of self-reference. The painter refers (iconically) to
the (his) hand, which he uses in painting.
The interesting and new process is No. 2; i.e. a vector-field prescribes the path of
the shaping energy from the right hand (R) to the left hand (L). If we magnify the zone
of contact, we see the bouncing of the tool on the zone. It has two effects:
It creates a hole at the point of contact.
It triggers a shock wave, which may split the zone.
Archaeologists can recognize the goal-directed activity of hominids (humans) by the
small hole; the intended effect is the splitting and a specific result in certain materials
is the sharp edge of the tool.4
From the pragmatic sequence, we can derive an idealized cognitive model of
events (cf. Lakoff 1987:6876), which is based on the body schema of human hands
and their instrumental use. One can distinguish simple and complex (interactive) submodels. The first four sub-models may be called simple:
1. Simple events are linked to one (left or right) hand.
2. Simple events involve some body acting on an individual object (bone, stone, etc.).
3. Simple events are modular insofar as they may be repeated, inserted into other contexts and combined with other events (self-containment is provided by the body).
4. The transmission of force is prototypically asymmetric. One hand (and a tool in this
hand) moves, the other hand fixes the object which has to be shaped. As the shaping
instrument is deformed to a lesser degree, an asymmetric effect is produced.

. The topic intention of knappers is critically discussed by Davidson (2002). He prefers to

assume intentionality only for Middle Stone Age knappers. Under this assumption, stone industries
would only be an indication of language capacity some 10.000 y BP. As a consequence, language,
like other technologies, would be a fact of cultural evolution rather than of biological evolution.
This conclusion is contradicted by other facts summarized in Wildgen (2004: Chapters2 and 4)
and would not explain the centrality of language for all existing human populations and the stability
of its acquisition.

Evolutionary pragmatics

The simple causal model is insufficient at the level of the basic instrumental action:
shaping a stone to obtain a hand-axe (or the flints, which are cut off; cf. Davidson &
Noble 1993). One has to consider an interaction between different types of causation:
5. An agent perceives/experiences affordances centered in the objects (cf. Gibson
1966); they have to be respected or exploited.
6. The cooperation of hand and eye (acting and perceiving) is strengthened in an
adaptive cycle.
7. The cooperation of right and left hand, of thumb and fingers, is further elaborated.
The simple billiard-ball schema of linear transmission of momentum fails as a causal
model. One has to define a concept of causation, which includes:
cooperation of body and environment, body-centre (e.g. brain) and periphery
(limbs, e.g. hands);
nonlinear-causation, i.e. catastrophic effects after the accumulation of minor
causes; and,
the branching (or diffusion) of effects.
The pragmatics of action with hands establishes a micro-level of emerging pragmatic
functions which elaborate the relation between cause and effect. At the macro-level,
human housing and house-building is a domain where structures emerge which can
be reorganized in human linguistic communication in the shape of space-oriented
communication, linguistic orientation in space, and memory of narrative contents
related to space.
3.3 Cro-Magnon life space and the pragmatic space of decorated caves
The term life-space as denoting the basis of human cognition was introduced by Kurt
Lewin, who observed the quickly changing perception and interpretation of space in
World War I (cf. Wildgen 2001). Life-space or cognitive ecology refers to the relevance
pattern, the meaning given to aspects of the surrounding space insofar as it is cognitively marked as a memory-system for what we have lived through, experienced,
enacted, imagined, hoped, and feared. These contents are attributed to spatial characteristics in a natural way. If in the first step of this process, real places receive memory
traces, in a second step the memory-space becomes purely internal and an artificial
(cognitive) space is constructed to receive and elaborate the mnemonic structure (cf.for
the art of memory, Yates 1966; Wildgen 1998).

93

94

Wolfgang Wildgen

Let us first consider the evolution of objective spaces used for memory traces
and then consider more abstract construed spaces. If we consider the life-space of
Cro-Magnon hunters, two regions are most relevant:
1. The space of hunting; it consists of the habitat, the migration routes of bison,
aurochs, reindeer, etc., the caves of bears and lions, the rivers rich in fish, etc. Together
with this hunting space, the sky with the motion of sun, moon and stars was probably
pragmatically organized as a memory-system of spatial orientation (B1).
2. The space of shelters, abris, cave openings, where the clans stayed for certain
periods of the year (B2).
The two base-spaces B1 and B2 subdivide social life in an external (open) and an
internal (closed) one, and may be blended or transformed in ritual, religious contexts.
Thus, the space of the sacred, magical, and ritual is one derivation, the space of burial
and life after death another one. This allows us to state three major trends:
1. The space for rituals and magic is derived from B1 and B2. Thus, the painted
caves are a derivation of decorated abris, cave entrances, by their transfer into dark
and hidden (normally not accessible) caves. We call this transferred space, the
ritual space (R).
2. The space for burials was in most cases not in closed caves, but rather in open
space. Nevertheless, these places could be blended with space R, e.g. in Neolithic
dolmens an artificial closed space covered with soil is placed in open space but
construed as a closed space. The Egyptian mastabas and pyramids correspond
topologically to this type (open, visible architecture with a hidden cave inside);
the burial caves in the Valley of Kings in Egypt are also of the same type, as the
mountain above was considered as a natural pyramid.
3. The internal structure of the natural and the construed caves has topologically
(i.e.ignoring all the topographical details) the shape of a closed tunnel, which may
be broken up by sub-tunnels.
One could consider further blends. A cave is like the inner space of the body: mouth
(nose) stomach intestines; or it is a negative of the body itself with head (entry)
neck (narrow entry) trunk (main room) limbs (side-rooms).
One could venture the hypothesis that the topology of life-space and body is the
stable background of semiosis. The (catastrophic) transitions to reinterpretations in
other (homologous) spaces constitute the proper semiosis beyond perceptual categorization. This corresponds to Peirces concept of a symbol created by transfer from one
sign-system to another (cf. Peirce 2002 [1865]: 105106). The regress of further and
further transitions may be controlled by topological invariants or by rather concrete,
iconic signs like the representation of animals, which probably have meanings in a

Evolutionary pragmatics

sign system beyond a description of contemporary fauna, but are anchored in visual
experience (contrary to abstract signs which accompany them).
4. From ecological to cultural pragmatics
The background of these processes is given by the ecological/situational context. Some
objects or context features become culturally significant. These are mainly:
a.
b.
c.
d.
e.

places (of living, of chase, etc.),

tools and the techniques of their use,
motion patterns, gestures, gestured signs, dance,
sound patterns (music, language),
art (engraving, colored surfaces or bodies).

The relevance of places (in space and time), of spatial orientation and categorization
are of primordial importance for the semantics of natural languages as the tradition
of localistic theories shows (cf. Wildgen 1985: Chapter 1 for a review of localistic case
theories). The relational structure of spatial networks is a precondition for migration patterns, contact with neighbors (trade exchange) and the local stratification of
human habitats. The micro-spatial categorization is relevant for goal-oriented manual
activities like throwing, hitting, but it is also the basis for the manufacturing of tools
(cf.Wildgen 2004: Chapter 4). Bodily motion in space and the coordination of spatial
locomotion in a group of individuals is the precondition for cooperative chasing and
other types of cooperative activities. Finally, external, perceivable processes become
the recipients of internal content (in dance, music, gestures, and language).
The cognitive and the social route enter a cycle of coordination, which tends to
induce individuals to select cultural contents as cognitive contents and to eliminate a lot
of potential contents which are not socially relevant. This strongly selective cycle may
be called socio-cognitive. In the two periods in which new behavior surfaced (atthe
stage of Homo Erectus and of Homo Sapiens), a dramatic co-evolution and selection
in the socio-cognitive evolution must have occurred, which has selected humans for
symbolic competence. In the co-evolutionary system between a cognizable ecology
and cognition/symbolic behavior it seems difficult to find clearly separated levels of
increasing complexity. Therefore, the following hierarchy is rather tentative:
a. Already in the last common ancestor of humans and chimpanzees (LCA), contextual space acts as an external memory of affordances, which is indexically
given by paths (of social locomotion and predator/prey-locomotion), harvesting
locations (and times), dangerous locations, places for sleep, courtship, housing,
frontiers of territories, etc. These indexically loaded areas and places function like
a catalyst of social action, insofar as they can coordinate social perception and

95

96

Wolfgang Wildgen

action. Their catalytic nature resides in their substantial independence, as they are
just a cognized reaction to the overall ecology but do not modify it.
b. As soon as space is more specifically organized in relation to cognition and social
use, it unfolds in a cycle of social investment. Architecture and the spatial organization of a village (or later a town) are clear examples. This level is autocatalytic
insofar as the spatial organization becomes itself a cyclic structure in which different functions cooperate. Figure3 sketches such an autocatalytic cycle (cf.Wildgen,
2008 and 2009b).

housing
fire place
myth. space

public space

ritual

tool making
outside
chase, harvest

Figure 3. Semiotic/functional subspaces and possible path-ways which link one to the other

In each subspace, specific symbolic media are rooted and co-evolve with them.
Thus the Paleolithic cave (in the Franco-Cantabric culture) is a specification of the
mythical/ritual space but is connected by its illusionist paintings to the outside space
of hunting. The relation is iconic, indexical (in its magical impact) and symbolic (in its
abstract signs; cf. Wildgen 2004:8083).
In subsequent development, when symbolic modes (e.g. languages and myths) of
different populations clash, e.g. in the large Neolithic societies of Egypt and Mesopotamia,
a new level of symbolic consciousness is reached. The single fields in Figure3 reorganize
in a hypercycle which produces a new, institutionalized symbolic system, e.g. a codified religion and a written language. Possibly the Franco-Cantabric culture (35,000 to
15,000) and later the Sahara cultures had already reached this level. As the code of its
abstract signs cannot be deciphered, this hypothesis cannot be substantiated.
5. Conclusions
The pragmatics of modern languages concern the embedding of linguistic utterances
into contexts of use (speaker, hearer, situation, time, etc.) and the action patterns,

Evolutionary pragmatics

formed by linguistic utterances (i.e. speech acts, conversational sequences). In an

evolutionary perspective, these contexts of action (the ecology, the group structure)
become dominant, because language itself is only emerging step-by-step and reshaping, developing the earlier action patterns. At the same time, the social ecology (and
later the physical ecology) is dramatically changed by the effect of linguistic thinking
and communication.
The field of evolutionary pragmatics focuses on the forces which shaped human
language, given the pragmatic and contextual structure of a population which was at
the stage of transition to a symbolic culture. This phase probably took a long time,
i.e. 21,7 million years, and, therefore, the pragmatics underlying human language
rather than more conventional symbolic features were the major field of constitutive
forces giving rise to human language, and all the complexities found in current grammars and lexicons.

References
Albertazzi, L. (ed.) (2001). The Dawn of Cognitive Science. Early European Contributors. Kluwer.
Bax, M., B. van Heusden & W. Wildgen (eds.) (2004). Semiotic Evolution and the Dynamics of
Culture. Lang.
Becker, P.R. (1993). Werkzeuggebrauch im Tierreich. Wie Tiere hmmern, bohren, streichen. Hirzel.
Bernstein, Basil, 1962. Social Class, Linguistic Codes and Grammatical Elements, in: Language and
Speech 5: 221240.
Boesch, C. (1993). Aspects of Transmission of Tool-use in Wild Chimpanzees. In K.R. Gibson &
T.Ingold (eds.): 171183.
Boesch, C. & M. Tomasello (1998). Chimpanzee and Human Cultures. Current Anthropology
39(5):591614.
Bhler, K. (1965) [1934]. Sprachtheorie. Die Darstellungsfunktion der Sprache, 2nd ed. Fischer.
Buekens, F. (1995). Pragmatism. In Handbook of Pragmatics.
Cassirer, E. (1944). An Essay on Man. An Introduction to a Philosophy of Human Culture. Yale
University Press.
(1957) [19231929]. The Philosophy of Symbolic Forms, vol. 1 to 3. Yale University Press.
Darwin, C. (1969) [1872]. The Expression of the Emotions in Man and Animals. Culture et Civilisation.
Davidson, I. (2002). The Finished Artefact Fallacy: Acheulean Hand-axes and Language Origins.
In A. Wray (ed.) The Transition to Language: 180203. Oxford University Press.
Davidson, I. & W. Noble (1993). Tools and Language in Human Evolution. In K.R. Gibson & T.Ingold
(eds.): 363388.
Dunbar, R. (1997). Groups, Gossip, and the Evolution of Language. In A. Schmitt et al. (eds.) New
Aspects of Humans Ethology: 7790. Plenum Press.
(2002). Human Evolutionary Psychology. Palgrave Macmillan.
Fehr, E. & J. Henrich (2003). Is Strong Reciprocity a Maladaptation? On the Evolutionary Foundation
of Human Altruism. CESIFO Working Papers No. 859.
Fitch, W.T., M.D. Hauser & N. Chomsky (2005). The Evolution of the Language Faculty: Clarifications
and Implications. Cognition 97: 179210.

97

98

Wolfgang Wildgen
Gibson, J.J. (1966). The Senses as Perceptual Systems. Houghten Mifflin.
Gibson, K.R. & T. Ingold (eds.) (1993). Tools, Language and Cognition in Human Evolution. Cambridge
University Press.
Green, Richard E. & many others. 2010. A draft sequence of the Neandertal genome. Science
(7May2010) 328 (5979): 710722.
Habermas, J. (1982). Theorie des kommunikativen Handelns, vol. 2: Zur Kritik der funktionalistischen
Vernunft. Suhrkamp.
Hewes, G.W. (1977). Language Origin Theories. In D.M. Rumbaugh (ed.) Language Learning in
Chimpanzee. The Lana Project: 253. Academic Press.
Immelmann, K. (1979). Einfhrung in die Verhaltensforschung, 2nd ed. Parey.
Jackendoff, R. & S. Pinker (2005). The Nature of the Language Faculty and its Implications for the
Evolution of Language (Reply to Fitch, Hauser & Chomsky). Cognition 97: 211225.
Jakobson, R. (1981). Linguistics and Poetics. In R. Jakobson Selected Wrtitings III. Mouton.
Lakoff, G. (1987). Women, Fire, and Dangerous Things. What Categories Reveal about the Mind.
Chicago University Press.
stman, J.-O. (1988). Adaptation, Variability, and Effect. IPrA Working Document 3: 540.
Pape, H. (1999). Pragmatismus. In H.-J. Sandkhler (ed.) Enzyklopdie der Philosophie: 12971301.
Meiner.
Peirce, C.S. (2002) [1885]. Semiotische Schriften, vol. 1: 18651903. Suhrkamp.
Piveteau, J. (1991). La main et lhominisation. Masson.
Quiatt, D. & V. Reynolds (1993). Primate Behaviour. Information, Social Knowledge, and the Evolution
of Culture. Cambridge University Press.
Reynolds, P.C. (1983). Ape Constructional Ability and the Origin of Linguistic Structure. In
E.deGrolier (ed.) Glossogenetics. The Origin and Evolution of Language: 185200. Harwood
Academic Publishers.
Tembrok, G. (1977). Grundlagen des Tierverhaltens. Vieweg.
Thom, R. (1983). Mathematical Models of Morphogenesis. Horwood.
Tomasello, M. (1999). The Cultural Origins of Human Cognition. Harvard University Press.
(2001). Primate communication. Handbook of pragmatics. John Benjamins.
Verschueren, J. & F. Brisard (2002). Adaptability. Handbook of Pragmatics. John Benjamins.
Wildgen, W. (1985). Archetypensemantik. Grundlagen fr eine dynamische Semantik auf der Basis der
Katastrophentheorie. Narr.
(1998). Das kosmische Gedchtnis. Kosmologie, Semiotik und Gedchtnistheorie im Werke von
Giordano Bruno (15481600). Lang.
(1999). Hand und Auge. Eine Studie zur Reprsentation und Selbstreprsentation (kognitive und
semantische Aspekte). Bremen U.P. (download from the homepage: http://www.fb10.uni-bremen.
de/homepages/wildgen/pdf/handundauge.pdf).
(2001). Kurt Lewin and the Rise of Cognitive Sciences in Germany: Cassirer, Bhler,
Reichenbach. In L. Albertazzi (ed.): 299332.
(2004). The Evolution of Human Languages. Scenarios, Principles, and Cultural Dynamics.
John Benjamins.
(2008). Semiotic Hypercycles Driving the Evolution of Language, in: Axiomathes 18(1):91116.
(2009a). Metarepresentation, Self-organization and Self-reference in the Visual Arts, in:
Wolfgang Wildgen and Barend van Heusden (eds.) 2009. Meta-representation, Self-organization
and Art, Lang, Bern: 173199.
(2009b). Sketch of an Evolutionary Grammar Based on Comparative Biolinguistics. in:
Rska-Hardy, Louise S. und Eva M. Neumann-Held (eds.). Learning from Animals? Examining
the Nature of Human Uniqueness, Psychology Press, Hove and New York: 4559.

Evolutionary pragmatics

(in print 2010). Thoms Theory of saillance and prgnance and Modern Evolutionary
Linguistics in: Wolfgang Wildgen and Per Aage Brandt (eds.). Semiosis and Catastrophes. Ren
Thoms Semiotic Heritage, Lang, Bern.
Plmacher, Martina & Wolfgang Wildgen (eds.) 2009. Prgnanter Inhalt Prgnante Form. Thematic
volume of: Zeitschrift fr Semiotik, 31 (12), Stauffenburg, Tbingen.
Yates, F.A. (1966). The Art of Memory. Routledge & Kegan.

99

Historical linguistics
Louis Goossens

1. Introduction
Historical linguistics (HL) is the branch of linguistics that is concerned with language change, especially over longer stretches of time. The two major questions that
have occupied historical linguists can be formulated as follows. (i) How do languages
change? (establishing the facts) (ii) Why do languages change the way they do? (the
search for explanations). The answers that have been given, as all answers in linguistics,
are at best partial, and reflect the specific biases of the schools or frameworks that have
succeeded each other in the course of time.
HL intersects with several other linguistic disciplines. The synchronic study of
older language stages in all their aspects, for instance, has direct relevance to our
understanding of the diachronic dimensions of language; minimally, it provides us
with the materials from which HL may proceed. Similarly, there is an outspoken link
with genetic linguistics, which deals with the historical classification of languages in
terms of their relationships and ancestry. Preferably, this happens on the basis of written
documents (or inscriptions) from the past, but also, in absence of written evidence,
by way of comparative induction and deduction. The so-called comparative method,
for that matter, is closely associated with HL; it has been used with success in the
reconstruction of older language stages for which no or incomplete data have been
transmitted to us. Other areas of investigation that can teach us a lot about language
change are the study of language death and the study of pidginization and especially
of creolization (defined as the way in which a pidgin acquires the status of mother
tongue for a given linguistic community). And it is obvious that this list could be
expanded further.
2. HL in pre-generative work
The oldest school, and one that is still operative today, is the philological-historical
tradition, which has greatly contributed to most of the full-scale grammars of older
language stages, and which, as a result of the textual orientation of philologists, often
provides us with rich and well-founded observations on older language phenomena in
context. For the linguist with an explicit pragmatic concern, the philological tradition
offers an immense quarry to be explored systematically.

Historical linguistics

It is in the wake of this tradition that the big dictionaries on historical principles
have to be situated, the Oxford English Dictionary and the like, which to date are the
most valuable studies of the lexicon for whoever wants to do work in HL. It was also
philologists who initiated Indo-European studies, beginning with the presidential
address in 1786 to the Bengal Asiatic Society by the British orientalist William Jones,
who was the first to point out the affinity between most European languages and
Sanskrit and to postulate a common origin for them. This was followed up in the early
nineteenth century by the work of comparative philologists such as Rasmus Rask,
Franz Bopp, August Schleicher and Jakob Grimm.
Jakob Grimms discovery of the correspondences between the Germanic voiced
and voiceless plosives (b, d, g; p, t, k) and the voiceless fricatives (f, , ) on the one
hand, and their counterparts in Indo-European on the other (Grimms Law) was
later refined by others, most notably Karl Verner, to the point where sound changes
became the main interest of the Neogrammarians. This school dominated HL from the
eighteen-seventies until the beginning of the twentieth century with their (explanatory)
hypotheses that sound laws are independent of change at the grammatical level, and
would be found to have no exceptions, as long as due attention was paid to all the
phonetic factors that surround them.
These hypotheses have since turned out to be untenable, but the vigorous research
which was carried out by the followers of this school, especially in morphological and
sound reconstruction in the context of the study of Indo-European, led to spectacular
advances in that field. It gave rise to the development of the comparative method,
which became the primary tool in the establishment of genetic units and the reconstruction of proto-languages.
This resulted in such a widespread genetic model as Schleichers Indo-European
family tree (Schleicher 1871), where the tree model shows the derivational history of
the different Indo-European languages. The tree model was challenged by Johannes
Schmidt (1872), whose wave theory emphasizes the gradual spread of the features
that help in distinguishing between two different languages, and which found confirmation in the isogloss maps of dialectologists after him. Still, the two models can be
regarded as complementary: trees represent the splits, waves the actual spread of the
features (Anttila 1989).
With the advent of structuralism, the epicentre of linguistic activity moved away
from HL to more synchronically oriented language studies. Still, it should be remembered that the father of structuralism, Ferdinand de Saussure, started out with an important contribution to Indo-European studies, more particularly to what was to become
known later as the laryngeal theory, which is concerned with the comparative reconstruction of the Proto-Indo-European sound system. More generally, structuralism
brought new impulses to linguistics which would also affect HL considerably. The
introduction of phonology had an important effect on the study of sound changes.

101

102 Louis Goossens

Sounds were now considered in their relation to the phonological system to which
they belong; the attention paid to the paradigmatic dimension of developments in
the phoneme inventory (through so-called splits or mergers) gave new perspectives
to the study of sound changes. Members of the Prague School (Jakobson, Trubetzkoy)
formulated insights into the structure of the phoneme as analyzable in terms of bundles
of articulatory and acoustic distinctive features, which made it possible to state precisely the effect of a particular change on the internal structure of the system as a
whole, especially in the case of sound changes affecting whole series of sounds in a
parallel way. A notable example here is that of the so-called Great Vowel Shift in English,
which was described by Martinet (1955) in terms of push chains and drag chains in
which the same phonetic features were at work.
Although structuralism contributed mainly to the area of (morpho-)phonological
change in HL, it also produced new insights into semantic developments. The theory
of word fields, i.e. the view that the lexicon is structured with respect to given conceptual areas, had its impact on the exploration of semantic change. For example in
the work of Jost Trier (1931), who was among the first to show how the structure of a
semantic field developed over a period of several centuries.
Dialect studies, more particularly dialect geography, also made important contributions to HL. Already at the time when the Neogrammarians formulated their
regularity postulate, Georg Wenker started an investigation to establish the precise
geographical boundary between High and Low German, more particularly that resulting
from the so-called Second (or High German) Consonant Shift (the shift of p, t, k to
[pf], [ts], []). Wenker and subsequent researchers found that individual lexical items
within a phonologically defined group are affected one by one in no definable order.
This contradicted the idea of a uniform change and showed that there is a step-by-step
spread through the lexicon with possible co-existence of old and new pronunciations
at a given time; what we have here is the (partial) forerunner to the insights into variability associated with later sociolinguistic work.
Important investigations on lexical change by dialect geographers include that of
Jules Gilliron in his preparation of the Linguistic Atlas of France (19021912). Among
his discoveries the best known is probably that resulting from his dialect maps for the
words cat and cock (rooster) in South-western France: Latin gallus cock and cattus cat
fell together as gat due to a regular sound change; as a result we get other words for
cock, such as faisan pheasant. Gilliron showed, in other words, that a factor operative in constraining lexical change was the avoidance of homonymic clashes, the loss
and replacement of a lexical item by another when two or more formerly distinct items
undergo phonological changes that make them homonymous. This demonstrated the
possible interdependence of semantic and phonological developments.
Although dialect studies on (Indo-)European languages have diminished considerably in the totality of linguistic research undertaken these days, there has been

Historical linguistics

an enormous expansion as regards other language families; witness the ever-growing

body of work on, especially, Amerindian, Oceanic and African languages. Descriptive
studies of this type no doubt continue to date to make empirically well-founded contributions to our insight into language change.
A methodology relevant to HL which stands apart from the different frameworks
that successively dominate the linguistic scene, is lexicostatistics or glottochronology
as devised by Morris Swadesh and Robert Lees. Lexicostatistics is a technique which
compares different languages with respect to the words figuring in a so-called basic
word list; in deciding that two words are cognates, allowance is made for the effects
of phonetic change. The percentage of shared items is correlated with the minimum
number of centuries of divergence between the languages compared, which provides
us with a statistical methodology to compute the time at which related languages have
diverged. Glottochronology can be criticized for several reasons, including, for example,
the fact that our ability to establish a basic word list that would be suitable for all languages is an illusion. Still, the method has shown its usefulness in genetic linguistics
(see some of the applications reported in Baldi 1990). Among the attempts to come up
with new strategies to calculate how old languages are, we mention Nichols (1992).
3. HL and generative grammar
The Chomskyan revolution in linguistics had an important impact on HL as well. It
shifted the centre of interest even more to synchrony, at the expense of HL, though,
of course, work on language change was also affected considerably. The main concern
in the generative model has been with syntax (conceived of as the central component
of the grammar) or, in more recent terminology, the search for universal grammar.
In its first two decades, however, also with respect to HL, phonology, one of the two
interpretive components, received considerable attention as well. The same cannot be
said of the other interpretive component, the semantic one, whereas pragmatics was
expicitly banned as being outside the concerns of the model.
Generative phonology takes the output of the syntactic component, with inserted
lexical formatives specified as matrices of (universal) phonetic distinctive features,
as its input. These are converted by phonological rules to yield the required phonetic (often no more than a phomemic) output. As such, there is no longer place
for autonomous phonemes, and the morphophonemics of taxonomic grammars are
included in the scope of phonology. Phonological change is viewed as resulting from
differences between successive synchronic grammars which may be a matter of the
underlying forms of corresponding lexical items, as well as of transformational rules,
or the ordering of those rules. The changes are ascribed to the alterations that occur
when successive generations interpret the language they learn as their mother tongue.

103

104 Louis Goossens

Rule additions at the end of the phonological component may eventually lead to new
underlying representations, with ensuing simplification in the phonological rules.
Grimms Law, for example, is accounted for in these terms by Kiparsky (1965). Rule
reordering has been conceived as resulting in bleeding or feeding orders (the terminology is Kiparskys), according to whether the new ordering deprives a subsequent
rule of its input, or makes a new input available to it. Obviously, a lot depends on the
abstractness of the underlying forms adopted, for which there is quite some variation
in the different subcurrents of generative phonological thinking. A general weakness
is the tendency to proceed to generalizations that rest on insufficient empirical considerations of the data, and thus Chomskys requirement of empirical adequacy often
goes unheeded.
In a somewhat parallel fashion changes in the syntactic component may be situated either in the transformational rules (rule addition, simplification, reordering,
loss), or in the deep structure (the phrase structure rules, including the categories or
functions figuring in the base rules, which, again, have not been particularly stable
over the years). It is the latter changes that are most interesting from the point of
view of the search for new generalizations with respect to syntactic change. Lightfoots
work (1979, 1991) is a good example of the kind of explanation that typically emerges.
The recategorization of certain English verbs (including the modals) as AUX or, more
generally even, the parametric changes affecting the grammar of new generations
of language learners give us a broad view of language change in terms of universal
grammar as conceived by generative linguists. It would seem, however, that at present
the generalizations, interesting though they may be as hypotheses, are too sweeping to
stand the confrontation with detailed empirical work (as is carried out, for instance,
in most of the grammaticalization studies reported on in the next section). Quite
generally, they can at best offer very restricted explanations because they disregard
the social and pragmatic contexts in which all language use (and language change) is
embedded. Still, there is no doubt that generative grammar has given new impulses
in the search for explanations as to how and why languages change.

4. New perspectives
Whereas generative grammar promoted the search for explanations in HL, it also
imposes a kind of straitjacket which makes it impossible to include all the multi-varied
factors relevant to language change into the picture. A number of recent directions
in linguistic research have restored the broad horizon that is required for a full-scale
study of language change. In some respects this amounted to a continuation, or a
partial rediscovery, of older work. Successively we mention (i) continuing work in
genetic linguistics, (ii) the impact of socio-linguistics and of studies on creolization,

Historical linguistics

(iii)insights arising from a discourse perspective on language change, and (iv) studies in
grammaticalization (with a growing importance of a cognitive approach to linguistic
phenomena). The last three approaches place HL in an increasingly functional context
and are opening it up more and more to a full-scale pragmatic perspective.
i. Work in genetic linguistics has continued to attract linguists all over the world.
Interaction with work on language typology has resulted in comprehensive surveys of
the language families of the world, such as Ruhlen (1987). The comparative method
has been applied to other language families besides Indo-European, giving rise to
new insights as well as to new questions and problems (see again Baldi 1990).
Reliance on the comparative method has varying implications if the attestation
of older language stages is lacking, as is the case in most language families except for
the Indo-European, the Afro-Asiatic, and the Sino-Tibetan languages. In cases where
attestation is poor, increased interaction with typological studies becomes imperative
(as has been current for the American Indian languages). Even if attestation over long
stretches of time is available, the importance of the writing system which was used to
record language data may play a crucial role: ideographic or hieroglyphic scripts (as
were used for Chinese or Egyptian, for example) and to a lesser extent syllabic scripts,
do not reveal sound properties in the same way as alphabetic writing systems do.
As Dixon (1990) points out, the 200 to 250 aboriginal Australian languages are
particularly problematic with respect to the distinction between similarities that come
from diffusion (resulting from a good deal of multilingualism as a consequence of a
predominance of exogamous marriages, or from lexical replacements because of cultural taboos), and those that are due to genetic retention.
On the other hand, the complex morphology typical of, for example, a lot of
American Indian languages may provide new opportunities for the development
of reconstructive techniques and may lead to the sharpening of those techniques
(Campbell & Goddard 1990). Minimally work outside Indo-European has proved that
the comparative method works, while at the same time it has revealed some points on
which it has to be refined or improved.
ii. Whereas European dialectologists have always emphasized the variation in all components of language, synchronically as well as diachronically, the reaction against
the generative view of language change on the American scene is in the first place
associated with the work of William Labov. Beginning with his detailed studies of the
changes in the diphthongs [ay] and [aw] of speakers on Marthas Vineyard (Labov
1963), he demonstrated that sound change originates in a relatively small number of
lexical items, and that both its inception and generalization are correlated with social
factors. This resulted in a theory that views phonological rules as essentially variable,
i.e. as correlating with social variables, and that sees social motivation as a powerful
cause of linguistic change. This implies that sounds are regarded as sociolinguistically

105

106 Louis Goossens

marked, and that this variability is a motor of language change. Also, that it is not early
language acquisition that causes alterations in languages (the predominating hypothesis with generativists) but, rather, imitation of prestige models in peer groups.
Another area of investigation which is sociolinguistic in the broad sense of
the word and which is contributing important insights into language change, is
the study of pidgins and creoles. For one thing, it provides a serious challenge
to the traditional idea upheld by the family tree model that languages gradually
diverge over time from a common ancestor (see, e.g. Romaine 1994). For another, it
has offered far-reaching explanations about the evolution of language in the human
species, such as Bickertons (1981) bioprogram hypothesis, which claims that children in a pidgin-speaking community are confronted with a lack of adequate input
from the older generation and have to fall back on an alleged innate bioprogram
which contains a blueprint for the features the creole must have. Even if this hypothesis is controversial (to say the least), it nevertheless raises important questions with
respect to the role to be attributed to universalist explanations in our understanding
of language change.
iii. A lot of language change can only be understood if the broader context of language
use is taken into account. Syntactic changes affecting word order, for example, are as
a rule the result of the generalization of marked patterns which at first are used with
specific discourse functions. A well-known investigation working with this insight is
Vennemanns research (1974): he interprets the change in basic word order in English
from SXV to SVX as resulting from a discourse-functional pattern T(opic)VX, where
also the breakdown of Old English morphology and universalist tendencies to group
Operators and Operands are taken to have played an important part in the development. More generally, markedness shifts proposed in functional language models
(such as Dik 1989) deal with language change along these lines as well.
iv. Although grammaticalization studies have a long ancestry (the term grammaticalization was probably coined by Meillet in 1912, but the process had been noted by
other linguists before him), it has received renewed attention by a great many linguists
in recent times, with increasingly refined insights that hold promise for the future of
HL (Givn 1979; Bybee & Pagliuca 1985; Bybee & Dahl 1989; Traugott& Heine 1991;
Hopper & Traugott 1993). One important mechanism which has been explored in this
context is reanalysis, the change in the structure of (a class of) expressions that does
not involve any immediate modification of its surface manifestation (Langacker 1977);
the reasoning process relevant here is abduction, which plays an essential role in the
development of a great many linguistic patterns (see Andersen 1973).
It is through the cyclical interaction of reanalysis and analogy (to be understood
as the attraction of extant forms to already existing forms, i.e. some form of generalization) that new grammatical structures are created. Examples are the grammaticalization of English going to + infinitive from a verb of motion with an infinitive of
purpose to a marker of futurity, or the development of the negation (ne) pas in

Historical linguistics

French from the reinforcing element pas step, or the rise of new verb morphology
from analytical structures of various types.
An important idea put forward in this context (see Hopper & Traugott 1993:
Ch.57) is the hypothesis of unidirectionality. It works from the evidence that there
is a vast number of known instances of the development of grammatical structures or
formatives which involve the change of a lexical item or phrase through discourse use
into a grammatical item. In this development, a decategorialization takes place from
a major category (especially nouns or verbs) to a minor category (mainly adpositions
or auxiliaries) and possibly, but not necessarily, to affixes. The opposite directionality
is not observed (instances involving the lexicalization of grammatical elements, as in
German duzen use the familiar form of address from the familiar pronoun du, are
regarded as illustrating a general process of word formation that does not invalidate
the hypothesis). In its strongest form the unidirectionality hypothesis would claim that
all grammatical items are derived from lexical material.
Detailed studies in grammaticalization reveal that language change is to be understood not in terms of one specific cause or explanation but, rather, as pointed out
by Hopper & Traugott (1993), in terms of a multiplicity of motivations. A crucial
ingredient, besides language acquisition factors and all sorts of contacts within (or
among) linguistic communities, is the pragmatic negotiation of meaning in communicative situations, for which speakers and hearers depend on their cognitive processing
of language in use, is.
Among the cognitive processes thought to be relevant in grammaticalization, and
more generally as well, renewed attention is going to metaphor and metonymy (see for
example the interpretation of the modal verbs in English in these terms in Sweetser
1990), though, again, they had been recognized for a long time as powerful extension
mechanisms in lexical change (see, for example, Ullmann 1967). Generally, the insight
in cognitive linguistics that linguistic categories are inherently polysemous as well as
inherently dynamic is bound to be helpful in understanding language change.

5. Explicit
In conclusion, let us take a brief look at HL as it appears to present itself today. The
main point, it would seem to me, is that in more than one respect HL is broadening its
scope, which multiplies the tasks ahead but which should also result in a much better
understanding of its object.
While continuing its efforts in areas that have been its traditional concern, such
as in genetic linguistics, HL is now giving these studies a world-wide character. The
perspective at present goes well beyond Indo-European to include language families
that present new challenges with respect to the type of language data as well as regards
the methodology with which to proceed.

107

108 Louis Goossens

Perhaps more importantly, an increasing number of diachronic linguists now

embraces the view that languages are essentially heterogeneous and that the result of
language change can only be a different kind of heterogeneity. This view implies the
necessary integration of sociohistorical factors into HL and the recognition of language attitudes (for individuals) or, more collectively, ideologies, as (co-)determinants
of language change (see Thomason & Kaufman 1988 & Meeuwis & Brisard 1993).
More generally, this means that changes are regarded as the unintended collective
consequence of individual intended actions (see, for example, Keller 1991). All this is
contributing to as well as supported by a more precise insight into the causality that
underlies language change; a view which can be characterized as non-nomic or nondeterministic (see, for example, Itkonen 1983) and which is in line with the revival
of a hermeneutic, context-embedded interpretation of what is going on in languages
(cf. Anttila 1989).
The combined emphasis on heterogeneous dynamicity, on the pragmatic context
of all human language, and the renewed effort to lay bare the essentially flexible conceptualizations of linguistic symbolizations (most prominently in work like Langacker
1991) is bound to open up a lot of new insights (and to rediscover a great many old
ones) in HL, as in language studies in general.

References
Andersen, H. (1973). Abductive and deductive language change. Language 49: 765793.
Anttila, R. (1989). Historical and comparative linguistics. John Benjamins.
Baldi, P. (ed.) (1990). Linguistic change and reconstruction methodology. Mouton de Gruyter.
Bickerton, D. (1981). Roots of language. Karoma.
Bybee, J.L. & . Dahl (1989). The creation of tense and aspect systems in the languages of the world.
Studies in Language 13: 51103.
Bybee, J. & W. Pagliuca (1985). Cross-linguistic comparison and the development of grammatical
meaning. In J. Fisiak (ed.) Historical semantics and historical word formation: 5983. de Gruyter.
Bynon, T. (1977). Historical linguistics. Cambridge University Press.
Campbell, L. & Y. Goddard (1990). Summary report: American Indian languages and principles of
language change. In P. Baldi (ed.): 1732.
Chomsky, N. & M. Halle (1968). The sound pattern of English. Harper and Row.
Crystal, D. (1987). The Cambridge encyclopedia of language. Cambridge University Press.
Dik, S.C. (1989). The theory of functional grammar, Part1. Foris.
Dixon, R.W. (1990). Summary report: Linguistic change and reconstruction in the Australian languages. In P. Baldi (ed.): 393401.
Geeraerts, D. (1988). Cognitive grammar and the history of lexical semantics. In B. Rudzka-Ostyn
(ed.) Topics in cognitive linguistics: 647677. John Benjamins.
Gilliron, J.L. (19021910). Atlas linguistique de la France. Champion.
Givn, T. (1979). On understanding grammar. Academic Press.
Hock, H.H. (1991). Principles of historical linguistics. Mouton de Gruyter.

Historical linguistics 109

Hopper, P.J. & E.C. Traugott (1993). Grammaticalization. Cambridge University Press.
Itkonen, E. (1983). Causality in linguistic theory. Croom Helm & Indiana University Press.
Keller, R. (1991). Erklrungsadquatheit in Sprachtheorie und Sprachgeschichtsschreibung. In
D.Busse (ed.) Diachrone Semantik und Pragmatik: 117138. Niemeyer.
Kiparsky, P. (1965). Phonological change. Ph.D. Diss., MIT.
Labov, W. (1963). The social motivation of a sound change. Word 19: 273309.
Langacker, R. (1977). Syntactic reanalysis. In C. Li (ed.) Mechanisms of syntactic change: 57139.
University of Texas Press.
(1991). Foundations of cognitive grammar, vol. 2. Stanford University Press.
Lehmann, W.P. (1992). Historical linguistics. Routledge.
Lightfoot, D. (1979). Principles of diachronic syntax. Cambridge University Press.
(1991). How to set parameters. MIT Press.
Martinet, A. (1955). Economie des changements phontiques. Francke.
Meeuwis, M. & F. Brisard (1993). Time and the diagnosis of language change. Antwerp Papers in
Linguistics 72.
Meillet, A. (1958). Linguistique historique et linguistique gnrale. Champion.
Nichols, J. (1992). Linguistic diversity in space and time. University of Chicago Press.
Romaine, S. (1994). Language in society. Oxford University Press.
Ruhlen, M. (1987). A guide to the worlds languages. Stanford University Press.
Schleicher, A. (1871). Compendium der vergleichenden Grammatik der indo-germanischen Sprachen.
Bhlau.
Schmidt, J. (1872). Die Verwandtschaftsverhltnisse der indo-germanischen Sprachen. Bhlau.
Sweetser, E. (1990). From etymology to pragmatics. Cambridge University Press.
Thomason, S.G. & T. Kaufman (1988). Language contact, creolization, and genetic linguistics. University
of California Press.
Traugott, E.C. & B. Heine (eds.) (1991). Approaches to grammaticalization, 2 vols. John Benjamins.
Trier, J. (1931). Der deutsche Wortschatz im Sinnbezirk des Verstandes, vol. 1. Winter.
Ullmann, S. (1967). The principles of semantics. Blackwell & Jackson and Co.
Vennemann, T. (1974). Topics, subjects, and word order. In J. Anderson & C. Jones (eds.) Historical
linguistics: 339376. North-Holland.

Historical pragmatics
Andreas H. Jucker

1. Introduction
Historical pragmatics (HP) is a field of study that investigates pragmatic aspects in the
history of specific languages. It studies various aspects of language use at earlier stages
in the development of a language; it studies the diachronic development of language
use; and it studies the pragmatic motivations for language change.
HP has established itself as an independent field of inquiry since the mid-nineteennineties, and various conceptualizations of the field have been proposed. Jacobs &
Jucker (1995), for instance, distinguish between pragmaphilology and diachronic
pragmatics. Pragmaphilology, in this view, comprises studies of older stages of a language from a pragmatic perspective, while diachronic pragmatics studies the development of various elements of language use in the course of time. This second approach
is further subdivided into diachronic form-to-function mapping and diachronic function-to-form mapping depending on the starting point of the investigation. A formto-function mapping takes the formal properties of a linguistic element as a starting
point and studies its changing communicative functions, while a function-to-form
mapping takes a specific communicative function as a starting point and studies its
changing realizations.
Traugott (2004) in her overview of HP focuses on the pragmatic mechanisms of
language change. Against the background of early work by Geis & Zwicky (1971), who
argued that invited inferences can become part of the semantics of an expression, she
discusses the role of conversational maxims in semantic change, and in particular the
question whether different conversational maxims motivate different types of semantic
change and whether Grices distinction between particularized and generalized conversational implicatures can help us understand the process of semantic change.
Brinton (2001a), on the other hand, gives an overview of historical discourse
analysis, which may be seen as a different term for text-based HP. She distinguishes
between historical discourse analysis proper, diachronically oriented discourse analysis and discourse oriented historical linguistics. The distinction is based on different
methodological perspectives. Historical discourse analysis proper, which corresponds
to Jacobs & Juckers pragmaphilology, is largely synchronic and studies discourse
functions in texts at a particular historical stage in the development of a language.
Diachronically oriented discourse analysis studies the evolution of discourse functions, such as politeness or speech acts, in the course of time. And discourse oriented

Historical pragmatics

historical linguistics seeks to identify the pragmatic motivations for semantic change.
For more recent overviews of HP see Taavitsainen & Fitzmaurice 2007 and in particular
the recent Handbook of pragmatics edited by Jucker & Taavitsainen 2010.

2. Origins
Jucker, Fritz & Lebsanft (1999) trace the origins of HP in the study of the Romance
languages, German and English (see also Jucker 2000a; Schrott & Vlker 2005). In
the study of Romance languages, HP has its roots in a strong interest in the spoken
language including the history of the spoken language. Pioneering studies in the
narrower sense of HP are Lebsanfts (1988) study of the inventory, the forms and functions of Old French greeting formulas, and Schlieben-Langes (1979) sophisticated
analysis of the Old Provenal romance Flamenco on the basis of Grices cooperative
principle. Sll (1974) and Koch & Oesterreicher (1985; see also Koch 1999) have provided the framework of a medial clear-cut dichotomy between the phonic and the
graphic and the conceptual and scalar opposition between spoken and written.
In the field of German studies, HP has its sources in literary studies, in the history of spoken German and rhetoric, and in the historical extension of speech act
theory and the theory of language games. Noteworthy early studies are Ehrismanns
(1901/1904) and Metcalf s (1938) analyses of address terms in Medieval and Early
Modern German respectively, and Ltschers (1981) study of swear words and insults
in Swiss German.
English HP received major stimulations both from historical sociolinguistics and
from corpus linguistics. Early studies in English HP are Finkenstaedts (1963) analysis
of the you-thou distinction in Middle English and in Early Modern English. Brown &
Levinsons (1987 [first published in 1978]) politeness theory turned out to be a major
influence for English HP, e.g. in the work by Brown & Gilman (1989), who applied it
to four plays by Shakespeare. In the nineteen-nineties the development of computerreadable corpora gave not only historical linguistics but also HP a major impetus. The
Helsinki Corpus, a computer-readable stratified corpus of English texts ranging from
about 750 to 1710, for instance, made such works as Taavitsainens (1995) detailed
analysis of interjections in the history of the English language possible.

3. Data problems
For older stages of languages we have no direct access to the intuition of native speakers,
and except for the very recent past there are no recordings of spontaneous spoken
language. The available data consist exclusively of written texts. But this is no longer

111

112

Andreas H. Jucker

seen as an insurmountable problem, in particular if the distinction between spoken

and written language is not conceptualized as a dichotomy. As Koch & Oesterreicher
(1985; see also Koch 1999; Culpeper & Kyt 2010:11) have shown, we have to distinguish between the dichotomy of the graphic and the phonic realization of a text on
the one hand, and the scale between communicative immediacy and communicative
distance on the other. A text in the graphic code, i.e. a written text, is typical for situations of communicative distance. In a legal text, for instance, the author and the audience do not share a common physical context; the style is formal, the text is monologic
with prescribed topic development and without any spontaneity. A text in the phonic
code, i.e. a spoken text, is typical for situations of communicative immediacy, such
as face-to-face conversations in which the speaker and the audience share a physical
context; the style is informal, the text is dialogic with free topic development and it is
characterized by the spontaneity of the participants. But apart from such prototypical
cases of communicative distance in the graphic code and communicative immediacy
in the phonic code there are many intermediate cases. A funeral oration, for instance,
is in the phonic code, but it shows most of the features of communicative distance,
while an Internet chat interaction is in the graphic code but shows most of the features
of communicative immediacy.
In HP, therefore, researchers tend to analyze either data that are characterized by
communicative immediacy, such as private letters or diaries, or data that are clearly
related to forms of spoken language, such as plays or court records. Letters are particularly useful because we often have information about the demographics of both
the sender and the receiver and their relationship. This provides a social context for
a pragmatic analysis. Fitzmaurice (2002a), for instance, examines familiar letters in
seventeenth- and eighteenth-century English from a pragmatic perspective to account
for the ways in which writers communicate with their readers. The Corpus of Early
English Correspondence (CEEC), compiled at the University of Helsinki, has been
used extensively for research in HP and in historical sociolinguistics. In its full version
it covers the time from 1417 to 1680 and comprises 2.7 million words. A special issue
of the Journal of Historical Pragmatics provides several case studies (see Nevalainen
2004). Nevala (2004) uses the corpus for her analysis of address terms in Late Middle
and Early Modern English. For a recent volume of articles on Late Modern English
correspondence see Dossena & Tieken-Boon van Ostade (2008).
Plays are useful in spite of their fictional nature because they depict spoken interactions between speakers of different social classes in a vast range of everyday situations. Brown & Gilman (1989), for instance, used four of Shakespeares plays for their
classic study of you and thou in Early Modern English, and Busse (2002, 2003) analyzes
the interaction of nominal and pronominal address terms in the entire Shakespeare
corpus. Fitzmaurice (2002b) provides a reading of British comic drama from Ben Jonson

Historical pragmatics

in the seventeenth to Steven Berkoff in the twentieth century on the basis of Horns
reworking of Grices conversational maxims.
Court records depict real interactions, albeit highly regulated and often very
formal ones (see for instance Collins 2001, who analyzes the use of direct speech in
Medieval Russian trial transcripts). Another special issue of the Journal of Historical
Pragmatics provides case studies of courtroom interaction in Early Modern English,
of seventeenth-century Russian trial dossiers, of trial testimonies in Old Polish and of
the Salem Witchcraft trials (see Kryk-Kastovsky 2006).
Culpeper & Kyt (2010) use the Corpus of English Dialogues 15601760, which
comprises a large range of speech-related texts, such as trials proceedings, witness
depositions, plays and fiction, for a detailed account of speech-based linguistic features.
These features include cohesive devices, pragmatic markers and elements such as O, ha
or alas, which they call pragmatic noise.
4. Topics
Various topics have established themselves as particularly productive areas of study.
In the following I will briefly sketch some of the work that has been carried out on
discourse markers, speech acts, and politeness.
4.1 Discourse markers
Discourse marker research on current languages usually concentrates on establishing
usage conditions both in terms of the pragmatic function(s) of specific discourse
markers and in terms of the sociolinguistic stratification of their occurrence. Similar
synchronic investigations have also been carried out for particular markers at various
older stages in the development of a language. Lenker (2000), for instance, focuses on
the two Old English discourse markers solice and witodlice, which are both normally
translated as something like really, truly, certainly, verily. She argues that a traditional
semantic analysis is insufficient, and that they should be seen as text-structuring discourse markers in Old English narrative discourse. They appear to be semantically
bleached and syntactically detachable from the sentence. They are indicators of thematic discontinuities and thus work as episode boundary markers.
Jucker (2002) gives an overview of the most frequent discourse markers that were
current in Early Modern English. These were oh, why, well, pray/prithee, and marry.
He shows that the distribution of these markers is genre-specific and linked to orality.
They only occur in plays, in fictional writing and in trial records. Moreover, several
of them tend to co-occur with specific speech acts. They all have interpersonal functions, and they are semantically bleached. In some cases, however, some residue of

113

114

Andreas H. Jucker

the original semantic meaning is still discernible, for instance in the case of pray or
prithee, which are markers of a polite request and derive from the phrase I pray you
or I pray thee. Blakes (2002) focus is even narrower. In his grammar of Shakespeares
language he analyses the discourse markers why, what, well and I say as they are used
by Shakespeare. Why and what also function as interrogative pronouns, but as discourse markers they indicate possible agreement and understanding, and impatience
or irritation, respectively.
In addition, researchers have tried to trace the diachronic evolution of individual
markers. Researchers who trace the diachronic evolution of specific discourse markers typically start with a close analysis of the forms and functions of their markers.
Brinton (1996), for instance, analyses the distributions and functions of a range of
Old English, Middle English and Early Modern English discourse markers, or pragmatic markers as she prefers to call them, (e.g. hwt, gan, or anon) before she goes
on to trace the evolution or grammaticalization process of each of these markers.
The marker anon, for instance, developed from its original propositional meaning at
once, immediately (Brinton 1996:111) to textual and interpersonal meanings, e.g. as
a marker of attentiveness meaning roughly coming, at your service. But the development is complex and takes several parallel routes. It does not follow a unilinear course
of development.
Onodera (2004) also starts with a synchronic analysis. She illustrates the usage
and functions of the discourse markers demo and dakedo in Present-day Japanese
before she goes on to trace the evolution of the form and meaning of these markers. In
Present-day Japanese these elements are markers of contrast, i.e. referential contrast,
pragmatically inferable contrast, functional contrast and contrastive actions (Onodera
2004: 83). Historically demo first appeared in the eighteenth century. It was always
used in unit-final position and had only an ideational meaning. When it started to
appear in utterance-initial position, it also adopted textual and expressive functions
and thus became a discourse marker. In Present-day Japanese all three functions still
exist (Onodera 2004:99).
Somolinos (2005) traces the evolution of the discourse marker voire in French
and Brinton (1998, 2001b, 2005) presents the histories of the discourse markers only,
look and (I) say in English. And Person (2009) compares the use of the various uses
of oh in Present-day English conversations to the use of oh and O in the Elizabethan
English of Shakespeare and finds that most of the modern uses are already attested in
Shakespeares writings.
4.2 Speech acts
The analysis of speech acts from a diachronic perspective has been very productive in
recent years in spite of thorny problems of methodology and data. Traditionally speech

Historical pragmatics

act research has relied almost entirely on the intuition of the researcher or on experiments with native speakers such as discourse completion tasks or role-plays and role
enactments. These methods are not applicable to older stages of the language. Instead,
researchers have to rely on corpus methods. They must try to locate specific speech
acts in available historical corpora, either by reading the texts in the old-fashioned
and very time-consuming way, which imposes severe limits on the amount of data
that can be studied, or by relying on corpus-linguistic methods of data retrieval. But
such methods depend on formulaic expressions or constructions that identify specific
speech acts. Deutschmann (2003:36) in his extensive corpus analysis of Present-day
British English apologies argues that apologies are accompanied by a limited set of
easily identifiable routine formulae. These formulae can be traced by corpus-linguistic
methods, but this ignores all the non-formulaic and idiosyncratic formulations.
Jucker & Taavitsainen (2000, 2008a) and Taavitsainen & Jucker (2007, 2008) argue
that historical speech act analysis can be viewed as a contrastive analysis across time.
Instead of comparing two different languages, different stages of the same language are
compared, except that in the historical analysis the two (or more) stages are ordered
into earlier and later stages, and that the later stages can be seen as (direct) developments of the earlier stages. As in a synchronic contrastive analysis, however, it is
essential to identify a tertium comparationis, i.e. an element that stays constant in the
analysis, such as the illocutionary force of a speech act. Thus, a diachronic analysis of
insults presupposes that Old English flyting, the insults of the fictional characters in
Chaucers Canterbury Tales, Shakespearian name-calling, the ritual insults of urban
African-American adolescents and the more recent Internet flaming are sufficiently
similar in their illocutionary force to warrant a comparison (see Jucker & Taavitsainen
2000). Such an assumption is not unproblematic. In fact, they argue on the basis of
the insults analyzed that not only the realizations of this speech act changes over time
but also the underlying speech function, and they propose a framework of a multidimensional pragmatic space in which specific realizations of insults are analyzed within
their context-specific, culture-specific and time-specific settings and in relation to
neighboring speech acts. They conclude that medieval flyting, Shakespearean namecalling and present-day flaming are not realizations of one and the same speech function of insult but they are different speech functions located in the pragmatic space of
antagonistic behavior (Jucker & Taavitsainen 2000:92). See also Ltscher (1981) on
insults and swear words in Swiss German, Lebsanft (1988) on greetings in Old French,
Schrott (2000) on questions in Old Spanish or Skewis (2003) on directive speech acts
in eighteenth-century Chinese.
Related to the problem of the identifiability of individual speech acts is the question
whether speech acts are performed directly or indirectly (see Bertuccelli Papi 2000),
and whether they are preformed explicitly or implicitly. Kohnen (2000a, 2000b, 2002)
provides extensive empirical evidence for his claim that in Old English directive speech

115

116

Andreas H. Jucker

acts tended to be realized as explicit performatives, while in Present-day English they

tend to be realized implicitly. (1) is a typical Old English example of an explicit performative realization of a directive with a verb of requesting in the first person indicative
active, an object denoting the addressee and a clause introduced by t describing the
desired action.
(1) Ic bidde eow t e ymon eowra sylfra, swa eowere bec eow wissia.
I ask you to take care of yourselves, as your books teach you.

(quoted after Kohnen 2000:304)

Arnovick (1999) focuses on the diachronic development of speech acts. She provides
seven case studies in English illocutionary development. These include flyting and
sounding, promises, curses, farewells and sneeze blessings. By tracing the histories
of these speech acts, Arnovick identifies a range of processes and trends, such as
increased epistemicity, subjectification, and discursization (a special kind of pragmaticalization). She shows, for instance, how the Present-day English sneeze blessing Bless
you! derived from a religious blessing in Old English. In Present-day English it can
have a continuum of functions from religious blessing and superstitious blessing to a
vague wish and a mere polite formula. A speaker who says Bless you! in response to
someone who sneezes may do this to wish the sneezer good health or as a mere polite
formula. The processes of subjectification and discursization lead from the original
illocutionary force of a religious blessing to the derived forces of a wish and a polite
formula, respectively. In the process of subjectification, the illocutionary force of a
wish may have derived from the illocutionary force of a blessing via the strengthening
of the speakers perspective through a pragmatic reanalysis. In the process of discursization, pragmatic meanings are derived from lexical material and these pragmatic
meanings are subsequently strengthened into a conversational routine (Arnovick
1999:131133).
Other speech acts that have recently been analyzed from a diachronic perspective
are promises (see Pakkala-Weckstrm 2008; Valkonen 2008) and compliments (see
Taavitsainen & Jucker 2008b; Jucker et al. 2008).
4.3 Politeness
Brown & Levinsons (1987) view of politeness as a redress to face-threatening acts has
been and probably still is the most influential approach in politeness research in spite
of criticism and alternative suggestions (see, for instance, Watts 1992, 2003). Brown&
Gilman (1989), in a seminal paper, analyze four plays by Shakespeare on the basis
of Brown & Levinsons notions of power, distance and ranking of a face-threatening act.
They provide a categorization of the various substrategies of positive and negative politeness (e.g. the positive politeness strategy Notice admirable qualities, possessions, etc.

Historical pragmatics

or the negative politeness strategy Be conventionally indirect, 1989:167, 168) and

of names and titles, such as honorific adjectives (valiant Othello), unadorned titles
(general) and adorned titles (good my lord). On this basis they score pairs of minimally contrasting dyads and assess whether they bear out the predictions of the theory.
Itturns out that the results for power and ranking are those predicted by the theory,
but the results for distance are not.
Kopytko (1993, 1995) also based his analysis of positive and negative politeness on
Shakespeares plays, where he found a striking predominance of positive over negative
politeness strategies, and he concluded that it may be tentatively proposed that the
interactional style or ethos of British society has evolved from the dominating positive politeness culture in the sixteenth century towards the modern negative politeness
culture (Kopytko 1995:531).
Kohnen (2008a, 2008b) argues that politeness in Old English was characterized
by the notions of mutual obligation and kin loyalty of Anglo-Saxon society and by
the added Christian values of caritas and humilitas. Face concerns, on the other hand,
played only a marginal role if any at all. Jucker (2010) adds that the Middle English
concept of curteisie courteous behavior provides the necessary link between the
Anglo-Saxon type of politeness and politeness in Early Modern England, when facebased forms of politeness started to play a more significant role (see also Jucker 2008
and on historical (im)politeness research in general the recent volume by Culpeper&
Kadar 2010).
A topic of particular interest within the research of politeness is the study of
terms of address. Many languages use different pronominal terms of address for single
addressees, such as tu and vous in French, tu and usted in Spanish, ty and vy in Russian
and so on (cf. Jucker & Taavitsainen 2003:4). As part of HP, researchers have analyzed
earlier systems and their diachronic developments (see the papers in Taavitsainen &
Jucker 2003).
In English the polite pronoun ye for a single addressee first appeared in the
thirteenthcentury and the difference between ye (later you) and thou persisted into
the seventeenth century when you slowly became the general form and ousted the
former pronoun thou. Finkenstaedt (1963) and Breuer (1983) are two important early
analyses of this distinction (for more recent approaches see, for instance, Hope 1993,
1994; Jucker 2000b and several papers in Taavitsainen & Jucker 2003). In spite of the
common origin and the many similarities between English and German, the situation in German is more complex (see Ehrismann 19011904; Metcalf 1938; Listen
1999; Simon 2003 and Hickey 2003). In Present-day German the third person plural
pronoun Sie is used as the polite pronoun of address, but in the history of German,
first the second person plural pronoun ihr was used for this purpose, the third person
singular pronouns er and sie and several nominal forms of address, such as der Herr
the gentleman or meine Schne my beautiful one were added to the system in the

117

118

Andreas H. Jucker

course of time before it was reduced again to the simple opposition between du (the
old second person singular pronoun) and Sie.
Several researchers have pointed out that pronominal terms of address should
be analyzed in their co-occurrence with nominal terms of address, for instance Busse
(2002, 2003) or Mazzon (2003), who both use plays by Shakespeare for their analysis. And Nevalainen & Raumolin-Brunberg (1995) present a very comprehensive
analysis of nominal forms of address in letter salutations in Late Middle and Early
Modern English.
5. New perspectives
In spite of some relevant work that goes back to the beginning of the twentieth century, and in spite of an impressive amount of work that has appeared over the last
fifteen years or so, HP is still a relatively young field. In terms of established research
methodologies and established facts it cannot yet compete with fields such as historical
semantics or historical syntax, or with fields such as speech act theory or discourse
analysis (cf.Jucker 2000a). HP has made important discoveries by extending pragmatic
research methods to the history of various languages. It has shown how historical data
can be used for rich and rewarding analyses. However, many areas are as yet unexplored.
In particular, the analyses should be extended to other languages. So far, work in HP
has been more or less restricted to English, to a few other European languages, such
as German, French and Spanish, and to Japanese. Work on other languages is almost
non-existent. But even for English our knowledge is very patchy, as the brief overviews
in the sections above have indicated. Some speech acts have been studied in detailed,
but even for these we can only highlight some of their uses at particular points in time.
We do not yet have the methods to trace a truly diachronic history of specific speech
acts in specific languages. We know even less about the entire inventory of speech acts
at particular points in time and how such inventories changed in the course of time.
Likewise we know something about specific types of discourse or specific genres
at particular points in the history of individual languages, their structures and their
communicative functions, but we know little about how such discourse types or genres
developed, how new ones came into being and old ones got lost, and we know little
about the entire inventory of discourse types and genres.
And finally we know very little about the larger communicative situation of earlier periods. We have some patchy knowledge for some languages of how scientific
knowledge, for instance, was transmitted from one generation to the other, or of how
current news was gathered and transmitted to a news-reading audience, but we do
not have the larger picture of how the dissemination of knowledge and information
developed in the course of time.

Historical pragmatics

References
Arnovick, L.K. (1999). Diachronic pragmatics. Seven case studies in English illocutionary development.
John Benjamins.
Bertuccelli Papi, M. (2000). Is a diachronic speech act theory possible? Journal of Historical Pragmatics
1.1: 5766.
Blake, N.F. (2002). A grammar of Shakespeares language. Palgrave.
Breuer, H. (1983). Titel und Anreden bei Shakespeare und in der Shakespearezeit. Anglia 101:4977.
Brinton, L.J. (1996). Pragmatic markers in English. Grammaticalization and discourse functions.
Mouton de Gruyter.
(1998). The flowers are lovely; only, they have no scent.: The evolution of a pragmatic marker
in English. In R. Borgmeier, H. Grabes & A.H. Jucker (eds.) Anglistentag 1997 Giessen. Proceedings:
933. Wissenschaftlicher Verlag.
(2001a). Historical discourse analysis. In D. Schiffrin, D. Tannen & H.E. Hamilton (eds.) The
handbook of discourse analysis: 138160. Blackwell.
(2001b). From matrix clause to pragmatic marker. The history of look-forms. Journal of
Historical Pragmatics 2(2): 177199.
(2005). Processes underlying the development of pragmatic markers: The case of (I) say. In
J.Skaffari, M. Peikola, R. Carroll, R. Hiltunen & B. Wrvik (eds.) Opening windows on texts and
discourses of the past: 279299. John Benjamins.
Brown, R. & A. Gilman (1989). Politeness theory and Shakespeares four major tragedies. Language
in Society 18(2): 159212.
Brown, P. & S.C. Levinson (1987). Politeness. Some universals in language usage. Cambridge
University Press.
Busse, U. (2002). Linguistic variation in the Shakespeare Corpus. Morpho-syntactic variability of second
person pronouns. John Benjamins.
(2003). The co-occurrence of nominal and pronominal address forms in the Shakespeare
Corpus: Who says thou or you to whom? In I. Taavitsainen & A.H. Jucker (eds.) Diachronic
perspectives on address term systems: 193221. John Benjamins.
Collins, D.E. (2001). Reanimated voices. Speech reporting in a historical-pragmatic perspective.
John Benjamins.
Culpeper, J. & M. Kyt (2010). Early Modern English dialogues. Spoken interaction as writing.
Cambridge University Press.
Culpeper, J. & D.Z. Kdr (eds.) (2010). Historical (im)politeness. Peter Lang.
Deutschmann, M. (2003). Apologising in British English. Institutionen fr moderna sprk, Ume
University.
Dossena, M. & I. Tieken-Boon van Ostade (eds.) (2008). Studies in Late Modern English correspondence.
Methodology and data. Peter Lang.
Ehrismann, G. (19011904). Duzen und Ihrzen im Mittelalter. Zeitschrift fr deutsche Wortforschung1,
1901, 117149; 2, 1902, 118159; 4, 1903, 210248; 5, 1904, 127220.
Finkenstaedt, Th. (1963). You and thou: Studien zur Anrede im Englischen. Walter de Gruyter.
Fitzmaurice, S.M. (2002a). The familiar letter in Early Modern English. A pragmatic approach.
John Benjamins.
(2002b). Plethoras of witty verbiage and heathen Greek: Ways of reading meaning in English
comic drama. Journal of Historical Pragmatics 3(1): 3160.
Geis, M. & A.M. Zwicky (1971). On invited inferences. Linguistic Inquiry 2: 561566.

119

120 Andreas H. Jucker

Hickey, R. (2003). The German address system: Binary and scalar at once. In I. Taavitsainen &
A.H.Jucker (eds.) Diachronic perspectives on address term systems: 401425. John Benjamins.
Hope, J. (1993). Second person singular pronouns in records of Early Modern spoken English.
Neuphilologische Mitteilungen 94: 83100.
(1994). The use of thou and you in Early Modern spoken English: Evidence from depositions
in the Durham ecclesiastical court records. In D. Kastovsky (ed.) Studies in Early Modern English:
141152. Mouton de Gruyter.
Jacobs, A. & A.H. Jucker (1995). The historical perspective in pragmatics. In A.H. Jucker (ed.) Historical
Pragmatics. Pragmatic Developments in the History of English: 333. John Benjamins.
Jucker, A.H. (2000a). English historical pragmatics: Problems of data and methodology. In
G.diMartino& M. Lima (eds.) English diachronic pragmatics: 1755. CUEN.
(2000b). Thou in the history of English: A case for historical semantics or pragmatics?
In Ch. Dalton-Puffer & N. Ritt (eds.) Words: Structure, meaning, function. A festschrift for Dieter
Kastovsky: 153163. Mouton de Gruyter.
(2002). Discourse markers in Early Modern English. In R. Watts & P. Trudgill (eds.) Alternative
Histories of English: 210230. Routledge.
(2008). Politeness in the history of English. In R. Dury, M. Gotti & M. Dossena (eds.) English
historical linguistics 2006. Volume II: Lexical and semantic change. Selected papers from the
Fourteenth International Conference on English Historical Linguistics (ICEHL 14), Bergamo,
2125August 2006:329. John Benjamins.
(2010). In curteisie was set ful muchel hir lest: Politeness in Middle English. In J. Culpeper&
D.Z. Kdr (eds.) Historical (im)politeness: 175200. Peter Lang.
Jucker, A.H., G. Fritz & F. Lebsanft (1999). Historical dialogue analysis: Roots and traditions in the
study of the Romance languages, German and English. In A.H. Jucker, G. Fritz & F. Lebsanft (eds.)
Historical dialogue analysis: 133. John Benjamins.
Jucker, A.H., G. Schneider, I. Taavitsainen & B. Breustedt (2008). Fishing for compliments: Precision
and recall in corpus-linguistic compliment research. In A.H. Jucker & I. Taavitsainen (eds.)
Speech acts in the history of English: 273294. John Benjamins.
Jucker, A.H. & I. Taavitsainen (2000). Diachronic speech act analysis: Insults from flyting to flaming.
Journal of Historical Pragmatics 1(1): 6795.
(2003). Diachronic perspectives on address term systems: Introduction. In I. Taavitsainen &
A.H. Jucker (eds.) Diachronic perspectives on address term systems: 125. John Benjamins.
(2008). Apologies in the history of English: Routinized and lexicalized expressions of
responsibility and regret. In A.H. Jucker & I. Taavitsainen (eds.) Speech acts in the history of
English: 229244. John Benjamins.
(eds.) (2010). Handbook of historical pragmatics. Mouton de Gruyter.
Koch, P. (1999). Court records and cartoons: Reflections of spontaneous dialogue in Early Romance
texts. In A.H. Jucker, G. Fritz & F. Lebsanft (eds.) Historical dialogue analysis: 399429.
JohnBenjamins.
Koch, P. & W. Oesterreicher (1985). Sprache der Nhe Sprache der Distanz: Mndlichkeit und
Schriftlichkeit im Spannungsfeld von Sprachtheorie und Sprachgeschichte. Romanistisches
Jahrbuch 36: 1543.
Kohnen, Th. (2000a). Explicit performatives in Old English: A corpus-based study of directives. Journal
of Historical Pragmatics 1(2): 301321.
(2000b). Corpora and speech acts: The study of performatives. In Ch. Mair & M. Hundt
(eds.) Corpus linguistics and linguistic theory. Papers from the twentieth international conference
on English language research on computerized corpora (ICAME 20) Freiburg im Breisgau
1999:177186. Rodopi.

Historical pragmatics

(2002). Towards a history of English directives. In A. Fischer, G. Tottie & H.M. Lehmann (eds.)
Text types and corpora. Studies in honour of Udo Fries: 165175. Gunter Narr.
(2008a). Linguistic politeness in Anglo-Saxon England? A study of Old English address terms.
Journal of Historical Pragmatics 9(1): 140158.
(2008b). Directives in Old English: Beyond politeness? In A.H. Jucker & I. Taavitsainen (eds.)
Speech acts in the history of English: 2744. John Benjamins.
Kopytko, R. (1993). Polite discourse in Shakespeares English. Wydawnictwo Naukowe Uniwersytetu
im. Adam Mickiewicza w Poznaniu.
(1995). Linguistic politeness strategies in Shakespeares plays. In A.H. Jucker (ed.) Historical
pragmatics. Pragmatic developments in the history of English: 515540. John Benjamins.
Kryk-Kastovsky, B. (2006). Historical courtroom discourse: Introduction. Journal of Historical
Pragmatics 7(2): 163179.
Lebsanft, F. (1988). Studien zu einer Linguistik des Grues. Sprache und Funktion der altfranzsischen
Gruformeln. Niemeyer.
Lenker, U. (2000). Solice and witodlice. Discourse markers in Old English. In O. Fischer, A.Rosenbach&
D. Stein (eds.) Pathways of change. Grammaticalization in English: 229249. John Benjamins.
Listen, P. (1999). The emergence of German polite Sie. Cognitive and sociolinguistic parameters.
Peter Lang.
Ltscher, A. (1981). Zur Sprachgeschichte des Fluchens und Beschimpfens im Schweizerdeutschen.
Zeitschrift fr Dialektologie und Linguistik 48: 145160.
Mazzon, G. (2003). Pronouns and nominal address in Shakespearean English: A socio-affective
marking system in transition. In I. Taavitsainen & A.H. Jucker (eds.) Diachronic perspectives on
address term systems: 223249. John Benjamins.
Metcalf, G.J. (1938). Forms of address in German (15001800). Washington University Studies.
Nevala, M. (2004). Address in Early English correspondence. Its forms and socio-pragmatic functions.
Socit Nophilologique.
Nevalainen, T. (2004). Letter writing: Introduction. Journal of Historical Pragmatics 5(2): 181191.
Nevalainen, T. & H. Raumolin-Brunberg (1995). Constraints on politeness: The pragmatics of
address formulae in Early English correspondence. In A.H. Jucker (ed.) Historical pragmatics.
Pragmatic developments in the history of English: 541601. John Benjamins.
Onodera, N.O. (2004). Japanese discourse markers. Synchronic and diachronic discourse analysis.
John Benjamins.
Pakkala-Weckstrm, M. (2008). No botmeles bihestes: Various ways of making binding promises
in Middle English. In A.H. Jucker & I. Taavitsainen (eds.) Speech acts in the history of English:
133162. John Benjamins.
Person, R.R., Jr. (2009). Oh in Shakespeare: A conversation analytic approach. Journal of Historical
Pragmatics 10(1): 84107.
Schlieben-Lange, B. (1979). Ai las Que planhs? Ein Versuch zur historischen Gesprchsanalyse am
Flamenca-Roman. Romanistische Zeitschrift fr Literaturgeschichte 3: 130.
Schrott, A. (2000). Qu los podri contar? Interrogative acts in the Cantar de mio Cid. Some examples
from Old Spanish on asking questions. Journal of Historical Pragmatics 1(2): 263299.
Schrott, A. & H. Vlker (2005). Historische Pragmatik und historische Variettenlinguistik. Traditionen,
Methoden und Modelle in der Romanistik. In A. Schrott & H. Vlker (eds.) Historische Pragmatik
und historische Variettenlinguistik in den romanischen Sprachen: 122. Universittsverlag.
Simon, H.J. (2003). Fr eine grammatische Kategorie Respekt im Deutschen. Synchronie, Diachronie
und Typologie der deutschen Anredepronomina. Niemeyer.
Skewis, M. (2003). Mitigated directness in Honglou meng: Directive speech acts and politeness in
eighteenth century Chinese. Journal of Pragmatics 35: 161189.

121

122

Andreas H. Jucker
Sll, L. (1974). Gesprochenes und geschriebenes Franzsisch. Erich Schmidt [3rd edition 1985, revised
and amplified by F.J. Hausmann].
Somolinos, A.R. (2005). From certainty to doubt: The evolution of the discourse marker voire in
French. In J. Skaffari, M. Peikola, R. Carroll, R. Hiltunen & B. Wrvik (eds.) Opening windows
on texts and discourses of the past: 301317. John Benjamins.
Taavitsainen, I. (1995). Interjections in Early Modern English: From imitation of spoken to conventions
of written language. In A.H. Jucker (ed.) Historical pragmatics. Pragmatic developments in the
history of English: 439465. Benjamins.
Taavitsainen, I. & S.M. Fitzmaurice (2007). Historical pragmatics: What it is and how to do it. In
S.M. Fitzmaurice & I. Taavitsainen (eds.) Methodological issues in historical pragmatics: 1136.
Mouton de Gruyter.
Taavitsainen, I. & A.H. Jucker (eds.) (2003). Diachronic perspectives on address term systems. John
Benjamins.
(2007). Speech acts and speech act verbs in the history of English. In S. Fitzmaurice &
I.Taavitsainen (eds.) Methodological issues in historical pragmatics: 107138. Mouton de Gruyter.
(2008a). Speech acts now and then: Towards a pragmatic history of English. In: A.H. Jucker &
I. Taavitsainen (eds.) Speech acts in the history of English: 123. John Benjamins.
(2008b). Methinks you seem more beautiful than ever: Compliments and gender in the
history of English. In A.H. Jucker & I. Taavitsainen (eds.) Speech acts in the history of English:
195228. John Benjamins.
Traugott, E.C. (2004). Historical pragmatics. In L.R. Horn & G. Ward (eds.) The handbook of pragmatics:
538561. Blackwell.
Valkonen, P. (2008). Showing a little promise: Identifying and retrieving explicit illocutionary acts
from a corpus of written prose. In A.H. Jucker & I. Taavitsainen (eds.) Speech acts in the history
of English: 247272. John Benjamins.
Watts, R.J. (1992). Linguistic politeness and politic verbal behaviour: Reconsidering claims for
universality. In R.J. Watts, S. Ide & K. Ehlich (eds.) Politeness in language. Studies in its history,
theory and practice: 4370. Mouton de Gruyter.
(2003). Politeness. Cambridge University Press.

Implicature and language change

Kate Kearns

1. Introduction
The basic components of semantic change are the two processes represented in (1).
(1) a. Fa > Fab Form F with sense a acquires an additional sense b
b. Fab > Fb Form F with senses a and b loses sense a

Process (b) must operate on a polysemous form. If we assume that polysemy must arise
by process (a), process (b) is dependent on process (a), but not vice versa. Accordingly,
the process of sense gain, as in (a), is the sine qua non of semantic change.
This paper reviews proposals that in some semantic changes, implicature is the
mechanism by which sense b initially comes to be associated with Fa. The establishment of b as a sense of F is attributed to conventionalization of the implicature.
The notion of implicature is not precisely defined, and may be compared to a
concept with prototypical and peripheral instances. A prototypical implicature is a
particularized conversational implicature as first proposed by Grice (1975), in which
the implicature is intended by the speaker, dependent on the particular context of
utterance, and calculated by identifiable inferential steps, including certain communicative principles as premisses in Grices theory, the Cooperative Principle and its
constitutive maxims, most importantly the maxims of Quantity and Relation. (Implicatures calculated under Quantity and Relation are discussed in more detail below.)
Implicatures of this kind, though differently classified, remain important in postGricean and Neo-Gricean theories of communication.
Grice also identified the more peripheral class of conventional implicatures, nontruth conditional inferences which attach to particular words by convention. A conventional implicature is not calculated or inferred under the Cooperative Principle
and maxims, and being attached to an expression by convention, cannot be detached
from it. If an expression E carries a conventional implicature I, any use of E will
carryI, but if the utterance can be paraphrased without E, I will not be present. Grices
example of this is the connective but, which he describes as an expression of conjunction carrying a conventional implicature of some sort of contrast between the two
conjuncts, making their conjunction unexpected. The implicature of but, illustrated
in She was poor but honest, disappears in a paraphrase without but as in She was

124 Kate Kearns

poor and honest. In short, a conventional implicature is noncalculable and detachable.

A conversational implicature, on the other hand, is detachable because (Manner implicatures aside) it arises from the sense of an utterance and not from its form. Any good
paraphrase of an utterance with conversational implicature I will still carry I.
This contrast presents an apparent tension for canonical conversational implicature as the mechanism underlying a sense gain Fa > Fab. Conventional implicature is
attached to a form, but it does not arise, being simply attached to the form as a listed
property, corresponding only to the stage Fab. A conversational implicature arises but
is not regularly attached to a particular form, giving only the process a > ab.
Bridging the divide between conventional implicature and particularized conversational implicature, and presenting as a candidate for the mechanism underlying
Fa> Fab, is Grices generalized conversational implicature. Generalized conversational
implicature is calculated under the Cooperative Principle, and thus arises. In addition,
it is relatively independent of particular utterance contexts by virtue of being fairly
regularly attached to particular expressions, as in, for example, Horns (1984) account
of scalar implicature involving expressions of quantity or degree ranked on a scale of
informational strength.
The identification of conversational implicature along Gricean lines as the central
prototype is consistent with the emphasis generally placed on this kind of implicature
in the literature. But as Levinson (1983:127) points out, Grice in fact intended the
term implicature to be a general cover term, to stand in contrast to what is said or
expressed by the truth conditions of expressions, and to include all the kinds of pragmatic (non-truth-conditional) inference discernible. Understanding implicature in
this broader sense shifts the focus away from the implicating speaker and towards
any means at all by which the hearer may draw inferences, whether or not they are
primarily grounded in communicative strategies. Here implicature may be taken to
include cognitively-based inferences, arising out of the structure of lexical concepts.
The next section reviews selected analyses in which particular conversational
principles are cited in instances of generalized conversational implicature initiating
semantic change. Distinguishing implicature and metaphor in processes of change is
addressed in Section3, and the notion of implicature emerging from this discussion is
clarified in Section4. Section5 outlines semantic change triggered by inferences stemming from lexical concepts, and a proposal for resolving the apparent conflict between
metaphor and implicature.

2. Analyses citing conversational principles

Specific analyses of conversational implicature in semantic change appeal not to Grices
original four maxims, but to the NeoGricean theories of Horn (1984) and Atlas &
Levinson (1981), which are briefly outlined here.

Implicature and language change

Grices (1975: 45) Maxim of Quantity has two clauses with opposing effects:
Clause1 Make your contribution as informative as is required (for the current purposes of the exchange), and Clause 2 Do not make your contribution more informative
than is required.
Horn compares this opposition to the general opposing forces in natural systems
which produce a balance between economy of effort and efficiency of output. He proposes two principles in replacement of Grices four: the Q Principle Make your contribution sufficient, corresponding to Clause 1 of Grices Quantity maxim, and the
R Principle Make your contribution necessary, corresponding to Clause 2 of Grices
Quantity maxim combined with Grices maxim of Relation or Relevance, Be relevant
(Grice 1975:46). Horn writes A speaker who says p may license the Q-inference
that he meant at most p; a speaker who says p may license the R-inference
that he meant more than p (1984:14). An example of Q-inference is the inference from It is possible that p to It is not likely or certain that p. An example of
R-inference is the inference from She was able to solve the problem to She solved the
problem. Generally the analyses reviewed here cite R-implicature.
Horn follows Grice in formulating his principles as guides to the speaker. Atlas&
Levinsons Principle of Informativeness is more directly focussed on the hearers interpretive calculation, assuming that it is a basic intuition that the information an
utterance gives an addressee depends in part on what he already knows, believes, presumes, or takes for granted, in short, on what is normally left unsaid (1981:40). The
hearer will understand an utterance by way of an inference to the best interpretation,
which according to the Principle of Informativeness is the most informative proposition among the competing interpretations that is consistent with the common ground
(1981:41), and best fits the communicative intentions attributable to the speaker
in light of what he has said (1981:42).
Both Horns R Principle and Atlas & Levinsons Principle of Informativeness license
inferences towards a more specific or more informative interpretation than the literal
content of the utterance. Both theories also emphasise the role in interpretive elaboration of general background assumptions, including normative scripts and stereotypes.
As Atlas & Levinson note, temporal, causal and teleological relations between events
are stereotypical in our common sense conceptual scheme (1981:42).
A well-known example of semantic change by implicature is the development of
causal meaning from purely temporal meaning with connectives such as since and
while. Geis & Zwicky (1971) offer the examples in (2) as illustrations of the common
inference from temporal sequence to cause. In each case, although the literal sense
of the sentence expresses only temporal sequence (or overlap in (2c)), there is a clear
inference that the first event causes the second.
(2) a. After a large meal, we slept soundly.
b. Having finished the manuscript, she fell into a swoon.
c. Martha observed the children at play and smiled with pleasure.

125

126 Kate Kearns

The emergence of causal since, as in Since I will be out of town tomorrow, I cant
use these theatre tickets, is attributed to the causal inference becoming a standard
sense of since. Geis & Zwicky do not class inferences of this kind, identified as the post
hoc ergo propter hoc fallacy, as conversational implicature in Grices theory. However,
Traugott & Knig (1991) attribute this inference to the Principle of Informativeness,
given that it creates an increase in informativeness. It can also be seen to draw on
stereotypical temporal and causal relations in our conceptual schema, according to
Atlas& Levinsons comment cited above.
The inference of causation is not confined to sequence, but also occurs with
temporal overlap, as Traugott & Knig point out with the examples in (3) (their (7),
1981:(197).
(3) a. I couldnt work when the television was on.
b. I cant sleep now that I am alone.

The conjunction while, originally OE pa hwile pe at the time that, expressed temporal
simultaneity. Developing into the late OE conjunction wile, it attracted inferences of
causation, as in (4), taken from Hopper & Traugott (1993:85).
(4) Thar mycht succed na female, Quhill foundyn mycht be ony male.
No female was able to succeed while any male could be found.
(1375, Barbours Bruce 1.60 [OED while 2a]

The causal inference with while did not become a standard sense in English, but
full semantic changes of the kind are found in the examples in (5), from Traugott &
Knig (1981:197).
(5)

OHG dia wila so so long as > German weil because

Latin dum when, as long as, because
French quand when, because
Finnish kun when, while, as, since, because
Estonian kuna while, as, since, because

Further semantic changes attributed to informativeness-augmenting inferences

include the development of concessives, and various meaning shifts which may be
broadly classed as increases in expressed subjectivity.
Traugott & Knig discuss the concessive expressions in (6) (their (12), 1991:(199),
which originated as expressions of simultaneity, concomitance or similarity.

(6) a.

connectives originally expressing simultaneity or temporal overlap:

while, still, yet; German zugleich, indes(sen), dennoch; French cependant;

Turkish iken while, although; Indonesian sekali-pun although (<at the same
time+even); Hawaiian oiai while, meanwhile, although; Japanese nagara
while, although

Implicature and language change

b. connectives originally expressing simple cooccurrence or concomitance, or

even similarity: EME withal (<along with the rest), all/just the same; German
gleichwohl, bei all; Danish evenwel; Rumanian cu toate ca with all (things) that;
Turkish bununla beraber together with this; Hopi naama-hin although
(<together + thus); Quileute -te with, although
c. negation of asymmetry (e.g. negation of opposition, negation of negative
correlation): notwithstanding, nevertheless, nonetheless; French nempeche que;
German nichtsdestoweniger

The development of a concessive inference with these expressions, subsequently an

established sense, is attributed to the need to establish the relevance of the stated
relation to the communication, given that the simple cooccurrence or concomitance
of two situations is generally not worthy of remark. Such cooccurrence is, however,
worth noting where it conflicts with normative assumptions about what kinds of situations naturally cooccur, as in Traugott & Knigs example He can play the Beethoven
sonatas and he is only seven years old, where the concessive inference is clear: Even
though he is only seven years old, he can play the Beethoven sonatas.
Although Traugott & Knig (1991) class the negative concessives in (6c) with the
others (as a kind of double negation being tantamount to a positive), Knig (1985:276)
indicates the likely involvement of the general pragmatics of negation: It is a wellknown fact that negatives are only uttered in a context where corresponding affirmatives have already been discussed, or else where the speaker assumes the hearers belief
in and thus familiarity with the corresponding affirmative. Citing the Maxim of
Quantity, Knig argues that, for example, an assertion of p nempeche que q with the
French connective nempeche que, literally not preventing, is not informative without
the additional inference if p then normally not-q. That is, the assertion that p does
not prevent q is not informative unless we assume that p normally would prevent q,
so assuming that the utterance is adequately informative, we infer (or perhaps accommodate the presupposition) that p normally prevents q.
Horn (1989:201) suggests a similar development for negation in general. He writes:
The extra implicature associated with negation [roughly not p implicates one might
believe or might have believed that p] does indeed derive from the prototype situation
in which a negative statement is less informative often infinitely less informative,
as in My favorite number is not 5 than its affirmative basis or common ground.
[] But the markedness of negation, born in the pure pragmatics of conversational
implicature, may tend to become partly conventionalized, with the result that all
negative statements [] are affected [] by this implicature.

Traugott (1989) presents a range of instances of semantic change which she identifies
as increases in expressed subjectivity, stemming from an informativeness-strengthening
inference, including developments in the English modals shall, must and will, speech act
verbs such as insist and suggest, and adverbs such as probably, apparently and evidently.

127

128 Kate Kearns

It is generally accepted that the epistemic meanings (that is, concerning knowledge,
belief and certainty) of must, shall (should) and will, illustrated in (7ac) below, are
later developments from earlier meanings, deontic (that is, concerning obligation and
permission) for must and shall and future for will, illustrated in (7df).
(7) a.

The house must be haunted.

The house is certainly haunted

b. They should be home by now.

They are probably home by now
c.

That will be our train on platform C.

That is sure to be our train on platform C

d. He must return at once.

He is required to return at once He is forbidden to not return at once
e.

He should return at once.

Accepted custom or authority strongly recommends that he return at once

f.

He will return at once.

His immediate return is predicted

The shift to epistemic meanings is taken to show an increase in expressed subjectivity

because epistemics are concerned with knowledge and belief states, and with first
person uses, the beliefs and conclusions of the speaker. In addition, the epistemic
meanings of these verbs first emerged as weakly subjective, expressing general probability and certainty, acquiring more strongly subjective uses over time, expressing
the speakers own assessment of probability or certainty. For example, will and would
expressing I conclude that , as illustrated in (8), did not appear in standard English
until the nineteenth century, and is still uncommon in American English.
(8) a. 1847 Ch. Bront, Jane Eyre XI: This will be your luggage, I suppose.
b. 1857 Mrs Gaskell. Life of Ch. Bront I IV 79: Of the younger ones I have
very slight recollections, save that one, a darling child, under five years of age,
was quite the pet nursling of the school. This would be Emily [i.e. this must
have been Emily]. (Traugott 1989:43, (28), (29))

Concerning speech act verbs such as insist and suggest, Traugott discusses developments
from directive meanings, comparable to deontic modality, to assertive meanings, comparable to epistemic modality. Directive and assertive meanings are illustrated in (9)
(Traugott 1989:44).
(9) a. I insist that he cheat in the exam tomorrow.
directive
b. I insist that he cheats, and has done so all semester. assertive

A similar development from permission to acknowledgment of a fact is shown for

allow in (10).

Implicature and language change

(10) I allow them to use the lower path as a shortcut.

permission
I allow that the committee was not briefed on this point. acknowledgment

In these developments, Traugott argues that the (later) assertive meaning is in a sense
epistemic and subjective, illustrating a shift to meanings expressing the speakers
belief state or attitude toward the proposition (1989: 44). I note in passing that
Traugott does not comment on the contribution of the first person subject to this
distinction: the directive/assertive contrast is also found in (11) below, but with a
third person subject (11b) does not express speaker attitude, and presumably cannot
be considered subjective.
(11) a. He insists that she cheat in the exam tomorrow.
b. He insists that she cheats and has done so all semester.

Traugott also discusses the development of epistemic meanings for adverbs such as
possibly, probably, evidently, apparently and obviously, all originating as adverbs of
manner modifying the verb. As with the epistemic modals, the epistemic meanings of
these adverbs were first only weakly subjective, later developing uses as clear expressions of speaker attitude. Examples of weak and strong subjectivity in epistemic evidently
and apparently are given in (12) from Traugott (1989:47).
(12) a. 1690 Locke, Hum. Und. II. xxix: No Idea, therefore, can be undistinguishable
from another for from all other, it is evidently different (evident to all, weak
subjective epistemic) (OED)
20th C?: He is evidently right (in the meaning I conclude that he is right;
strong subjective epistemic inviting the inference of some concession or doubt
on speakers part).
b. 1566 Knox Hist. Ref. Wks 1846 I 49: The Bischoppis hes had heirtofoir sick
authoritie upoun they subjectis, that appearandly thei war rather King, and thow
the subject.
The Bishops have heretofore had such authority over your subjects that to all
appearances they rather [than you] were the king and you the subject. (OED)
1846 J. Ryland, Fosters Life II 107: It has been remarked, and apparently with
truth (=I think, but do not vouch for it) (OED)

Observing that these developments increase coding of speaker informativeness about

his or her attitude (ibid.: 49), Traugott attributes the changes to the conventionalization of implicatures which strengthen relevance and informativeness, according to
the Principle of Relevance in Sperber & Wilsons (1986) Relevance Theory, Atlas &
Levinsons Principle of Informativeness, or Horns R Principle.
In the analyses above, Horns R Principle is cited as underlying conversational
implicature. Horn (1984) also claims that both the Q and R Principles play a part in
the semantic changes traditionally termed broadening and narrowing. It is not clear,

129

130 Kate Kearns

however, whether or not Horn considers these to be implicatures, as the Q and R principles in his system are overarching principles operating on language globally, and not
confined to implicature. Horn states that the tension between the Q and R principles
affects a wide range of phenomena, ranging from implicature and politeness strategies to the interpretation of pronouns and gaps, from lexical change to indirect speech
acts, from the interpretation of case marking in so-called split ergative languages to the
analysis of recorded conversational interaction, from the pragmatic strengthening of
apparent contradictory negation to the weakening effect of logical double negation
(1989:194). This passage suggests that Horn himself might not include lexical change
under implicature, even though his theory of the Q and R principles is well known as
a development of Grices theory of implicature. With this proviso, I include here a brief
account of Horns remarks on narrowing and broadening.
Horn suggests that narrowing generally involves an R-based shift from a set
denotation to a subset (or member) of that set, representing the salient or stereotypical
examplar of the general category (1984:32). The shift may be complete, as in Greek
alogon speechless one > horse, Latin fenum produce > hay, poison cognate with
potion, undertaker one who undertakes > mortician, and corn, used for the main
grain crop of the region, viz. wheat in England, oats in Scotland and maize in Australia
and the New World.
Alternatively, the original meaning may survive in contrast with the new narrower
meaning, presenting a pattern of autohyponymy the new sense of a term is a hyponym
of the original, more general sense. Horns examples of this include color for hue
excluding black, white and grey, as in color TV, temperature for fever, number for
integer, and drink for drink alcohol as in I dont drink. All these are classed as originally R-based inferences from a set to a salient or stereotypical subset.
Lexical narrowing leading to autohyponymy may also be Q-based. Here Horn
cites Kempsons (1980) discussion of the characteristic input structure, in which a
general term has a basic-level hyponym naming only one part of a two-part division of
the general terms extension: for example, the species term dog has the hyponym bitch
female dog, but no basic lexeme denoting male dogs. In a context where information
on gender would be relevant, use of the general term dog constitutes an avoidance of
the more informative term bitch, licensing the Q-inference not female and thus the
interpretation of dog as male dog. Horns examples of Q-based narrowing, where a
general term is understood in contrast with its one existing basic hyponym, include
cow excluding bulls, rectangle excluding squares, finger excluding thumbs, and gay
excluding lesbian.
Horn proposes that lexical broadening, shifting from a set to a superset, is always
R-based. Examples include Latin pecunia property or wealth in cattle (cf. pecu livestock, cattle) > wealth > money, the cognate Old English feoh cattle > property
(cf. Modern English fee), Latin adripare, arripare come to shore > French arriver

Implicature and language change

arrive, and Latin panarium breadbasket > French panier basket. Under R-based
broadening fall the generic uses of terms originating as names, such as xerox photocopy, vaseline petroleum jelly, hoover vacuum cleaner and kleenex paper tissue.
Conversational implicature is also considered to be involved in a number of developments in tense and aspect. A well-known example is the development of modal will
want, be willing > intend > prediction of future event. As Aijmer (1985:13) writes
of the first person use, If the speaker [of I will] is willing to do something, it follows
conversationally that he intends to do it and that the future action will take place. The
development of will is also traced in Bybee & Pagliuca (1987), along with a similar
development for shall owe > be right, commanded or decreed > intention (with
first person) > prediction. They write A statement of obligation by a first person
subject amounts to a promise to carry out the act, and, by implication, a statement
of intention, By the Middle English period, shall in first person expressions of
intention has become common, and approaches the sense of prediction (Bybee &
Pagliuca 1987:14).
Inferential changes are also at work in the development of perfectives and anteriors
with change of state predicates into predications of present states. The Kanuri examples
in (13) are taken from Bybee et al. (1994:288).
(13) Perfect suffix -n with dynamic predicates:
Md isn Modu has arrived
Sh rkn I have seen him
with change of state predicates:
Nngn
Rngn
Grgzn

I know
I am afraid
S/he is angry

The anterior of a change of state verb implies that the final state now holds, and the
hearer may infer that the present state is the main point of the communication, as in
The fruit has ripened implicating The fruit is now ripe and ready to eat.
3. Implicature and metaphor
Much of the discussion of implicature in semantic change has been in the context of
research on processes of grammaticalization. Many of the semantic changes attributed to implicature are part of grammaticalization processes, and a central issue in
the discussion has been the demarcation of domains of operation for implicature and
metaphor, long known to be a major mechanism in semantic change.
I note that a distinction between the two presupposes that metaphor is not
included in implicature, although an inferential basis for the interpretation of metaphor

131

132

Kate Kearns

is assumed by Grice, who proposes that metaphor is interpreted as a recognized flouting

of the Maxim of Quality/Truthfulness. Relevance Theory also analyses metaphor as
interpreted by conversational inference, and Warren (1992:74) notes that to understand a metaphor correctly the hearer must infer, commonly from context, the grounds
of comparison between the literal and novel referents of the metaphorical expression.
Implicature, then, is not to be so broadly construed as to include any inferred
content, but only inferences of certain kinds. The main criterion for distinguishing
implicature and metaphor recalls Grices characterization of metaphor as a flouting of
the Maxim of Quality. If I say of a bibliophile at a book market Leave him be hes a
pig in mud, the metaphorical interpretation blissfully content is not simply added,
but substituted for the literal content of pig in mud, which is false in this instance.
Generally, an expression in a given utterance is intended either in its literal sense or in
its metaphorical sense, but not both.
In contrast, a conversational implicature is added to the literal content of the utterance and both senses are intended together. Calculation of the implicature depends
on the assumption that the literal sense is also intended and offered as true (with the
exception in Grices theory and Relevance Theory of irony and metaphor). It follows
from this that if a semantic change is to be attributed to implicature, there must be
a stage in which the original meaning and the new meaning are both present in the
same contexts.
This point is raised by Bybee et al. (1994:200), arguing that the deontic to epistemic
developments of must and should came about by different mechanisms. They point out
that should can have both deontic and epistemic readings in combination with verb
phrases interpreted as having past, present or future time reference, as in (14).
(14) a. The letter should arrive sometime next week.
b. The letter should be in the mail.
c. The letter should have come last week.

This supports the view that deontic should gave rise to epistemic should by implicature.
The proposed inference is plausible with (14c), for example, if we know that the
sender was obliged to have the letter arrive last week, unless we know otherwise we
may infer that the letter probably arrived last week.
They argue that epistemic and deontic must, on the other hand, are in almost
mutually exclusive environments, assigning deontic must to future environments and
epistemic must to present and past environments, with the examples in (15).
(15) a. The letter must arrive sometime next week.
b. The letter must be in the mail.
c. The letter must have been in the mail.

Example (15c) can only be interpreted as past and has only an epistemic reading.

Implicature and language change

Bybee et al. claim that (15a) has only a deontic reading (The sender is obliged to
ensure that the letter arrives next week), but I note that an epistemic reading I conclude
that the letter will arrive next week may be available here, as also in That tree must fall
soon its badly diseased. I note also that a deontic to epistemic inference appears to
be available for (15a): from Someone is obliged to have the letter arrive next week one
may infer The letter will arrive next week. The inferred certainty here is epistemic, not
predictive, as the future interpretation comes from the adverbial (next week, soon) and
is shared by both senses of must.
The contrast in (15b), on the other hand, without a future-denoting adverbial, is
sharp. The deontic reading can only be future Whoever posts the letter is obliged to
have it in the mail at some future time, while the epistemic reading is present time The
letter is surely in the mail. Where deontic must triggers a futurative interpretation, as
in (15b), the likely inference is predictive rather than epistemic. This point is noted
by Traugott & Knig (1991:209) for She must be married, from which one may infer
She will be married. Traugott & Knig identify this inference as epistemic, but it is
more accurately an inference to prediction, as must is here the source of futurity. The
epistemic sense I conclude that she is married cannot be inferred from She must be
married understood as She is obliged to get married.
I note that the two senses of must also overlap in generic contexts such as The
floors must be washed thoroughly every day, and again the deontic sense may imply
the epistemic: if one knows that daily floor washing is required, one may infer that it
is carried out.
Although the claim that deontic and epistemic must have mutually exclusive
environments is too strong, these examples do support the claimed correlation between
shared context and plausibility of inference. Where both deontic and epistemic senses
of must share a temporal context, future or present generic, the deontic to epistemic
inference is available. But where the temporal contexts diverge, and deontic must itself
is the source of futurity, the main inference is predictive rather than epistemic.
The criterion of cooccurring senses raises another point concerning a number
of the examples mentioned above. Original senses commonly persist alongside later
senses. True polysemy is established when the later sense can occur in a context that
excludes the original sense, as demonstrated for since by Traugott & Knig (1991:194)
with the examples in (16).
(16) a. I have done quite a bit of writing since we last met. (temporal)
b. Since Susan left him, John has been very miserable. (temporal, causal)
c. Since you are not coming with me, I will have to go alone. (causal)

If the independent use of the later sense cannot be demonstrated, it may be that the
new sense is not a fully established sense of the form, but has stabilized as a generalized

133

134 Kate Kearns

implicature reliably attached to the form. This indeterminacy appears with concessive
while, as in (17).
(17) While I understand your difficulties, I cannot help you.

Although the concessive meaning is the main contribution of while in (17), the original meaning of cooccurrence of situations is still present. In short, the persistence of
co-distribution of interpretations not only indicates that the later interpretation originated as an implicature, but also suggests that it continues to have that status.

4. Implicature, metonymy and merger

The criterion of shared contexts discussed above is proposed as a test for the presence
of implicature in general, but strictly speaking it identifies a particular kind of implicature. As noted earlier, pragmatic theory gives inference a role in the interpretation of
figures of speech, including both metaphor and metonymy. Both of these are inferentially interpreted, working from the premiss that what the speaker meant is not what
she literally said, and the convention of truthfulness has been purposefully violated.
The inferred figurative sense is substituted for the literal sense of a figure of speech.
A more precise (and autohyponymous) sense of implicature has emerged in the
discussion on change, focussed on those implicatures which involve sense addition
rather than sense substitution. This class of implicatures is identified by the criterion
of shared contexts, influenced by Coates (1983) distinction between ambiguity and
merger. Where Fab represents ambiguity (or polysemy), senses a and b occur in different contexts, but in a merger, a and b occur together in the same contexts. Conversely, where Fab is recognized as a merger, the change is likely to be attributed to
implicature under the R Principle or the Principle of Informativeness primarily on the
grounds that information has been added, rather than by identifying the calculation
of the inference.
As indicated above, metonymy and metaphor are both instances of ambiguity,
not merger, and so metonymy does not belong with the narrower class of implicatures
identified here.
A rather different view is proposed by Traugott & Knig (1991:201203), who
outline a distinction between metaphor and implicature based on sense relations.
Ametaphorical relation is comparative and analogic, based on a perceived similarity
between the structures of the original concept and the novel concept, which is typically
more abstract. For example, terms denoting body parts frequently develop into terms
for spatial regions and relations, as in Yucatec pach back (body part) > behind, tan
front (body part) > in front of , and tsu marrow > in (from Heine 1997:37).

Implicature and language change

Implicature (narrowly construed), on the other hand, is based on non-comparative

association, likened by Traugott & Knig (ibid.:210213) to metonymy. Metonymy is
based on perceived or typical contiguity of the old and new referents. Central examples are based on physical contiguity, as in Latin coxa hip > French cuisse thigh, or
Sterns (1931) famous example of bede prayer > bead, based on the practice of counting the prayers of the rosary by counting the beads which represent them. Metonymy
may also be cited for instances of contiguity in the linguistic context, as in French
ne pas not a step > pas not, where negative meaning has transferred to pas from
its construction mate ne.
Traugott & Knig (1991:210213) propose that implicature is a kind of metonymy
in that the implicated meaning of an utterance is associated with the literal meaning
by being covertly present in the discourse context. The mechanism of implicature
signals the association or contiguity between the two meanings. Adopting the same
proposal, Hopper & Traugott (1993:81) write, we use the term [metonymy] in the
restrictive sense of conceptual association. This sense can be found in Stern, who, in
speaking of permutation, says it results from a word [being] used in a phrase where
a notion in some way connected with its meaning is liable to form an element of the
context (1931:353).
As stated above, implicature broadly construed includes both metaphor and
metonymy, and on the narrow construal of implicature based on the criterion of shared
contexts, sensitive to merger, both metaphor and metonymy are excluded. On this
view, contra Traugott & Knig, implicature is not subsumed under metonymy. Rather,
the relation between them discussed by Traugott & Knig is one of analogy.

5. Inferences based on lexical concepts

Conversational implicature is also appealed to in work by Bernd Heine and his associates on semantic changes which they attribute to inferences which highlight or select
in certain contexts particular components of a word sense or lexical concept. Several
authors (see, for example, Lipka (1985); Geeraerts (1997) and references cited there)
provide comparable accounts of semantic change by shifts in the status of elements of a
lexical prototype, although these analyses are not generally framed in terms of conversational implicature. As originally peripheral features become central and the originally
prototypical features become marginal, the meaning of a word gradually changes.
Heine et al. (1991) to some extent adopt Traugott & Knigs comparison between
implicature and metonymy. They construct a model of semantic change in grammaticalization in which the relation between metaphor and context-induced inference
(identified as a kind of implicature) echoes a previously established relation between
metaphor and metonymy.

135

136 Kate Kearns

Although metonymy and metaphor are based on distinct sense relations, many
metaphorical transfers can be seen to originate in metonymy. Taylor (1989), for example,
points out that the MORE = UP metaphor may have a metonymic basis in the image of
piling objects in a heap: as more items are added the top of the heap rises. The transfer
becomes fully metaphorical when it applies in domains where the metonymic link is
excluded, as in Prices are rising.
Heine et al. discuss a number of semantic changes showing the same pattern
accumulated shifts in meaning, initially inferential, may emerge as metaphorical shifts
when they cross conceptual domain boundaries. This is illustrated here with the
development of the Ewe noun v (human) child into a suffix with various senses in
different contexts. The cognitive concept CHILD contains, among others, the features
YOUNG, SMALL, and DESCENDANT OF, each of which may be the most salient
content of the word child in different contexts of use, as illustrated in (18).
(18) a.

b.
c.

If he was born in 1905, then he was still a child

when the war began.
This is a childs cello.
She is their only child.

YOUNG
SMALL
DESCENDANT OF

The Ewe examples below, taken from Heine et al. (1991:Ch.3), show how these inferrable peripheral senses give rise to new conventional meanings in particular contexts
of compounding or affixation.
(19) YOUNG

tsu
nyfnu
nyi
det

man
woman
cow
oilpalm tree

tsu-v
nyfnu-v
nyi-v
det-v

boy
girl
calf
young oilpalm tree

YOUNG > INEXPERIENCED

nfl

teacher

nfl-v

one who has just started teaching

INEXPERIENCED > APPRENTICE

dfyfl

healer

dfyfl-v

assistant or apprentice to healer

APPRENTICE > UNQUALIFIED

kul

driver

kul-v

driver who has not passed license test

UNQUALIFIED > UNSUCCESSFUL, WANNABE

kesinftf rich person

kesinftf-v parvenu; person who is not really rich

SMALL
xf
l
akpa

house
animal
fish

xf-v
l-v
akpa-v

small house, hut

small animal species
small fish species

Implicature and language change

SMALL > WEAK, INSIGNIFICANT

gbe
ya

voice
wind

gbe-v
ya-v

weak, faint voice

light wind, breeze

SMALL > COMPONENT, PART OF

nnono drinking nnono-v mouthful of liquid

Chains of meaning shifts such as child > young > inexperienced > apprentice >
unqualified > unsuccessful and child > small > component, part of are composed
of small inferential steps plausibly attributed to implicature. They can be classed as
metonymic inferences, so long as properties shared by a conceptual entity, or elements
coocurring in a conceptual schema, can be classed as contiguous.1 Such shifts also
appear to be metonymic in that sense substitution is involved for example, the -v
suffix meaning young or small is not required to denote a human, unlike the original
word. Heine et al. (1991) identify the individual steps in the chain as metonymic, but
apparently on the grounds, adopted from Traugott & Knig (1991), that implicature is
metonymic, and not on the grounds of substitution of contiguous senses. More distant
links in a chain such as child: unqualified or child: component, part of appear to be
analogically related and taken in isolation, present as metaphors. Accordingly, Heine
et al. propose that implicature/inference and metonymy operate at the microlevel of
change, producing over time what they term emerging metaphors at the macrolevel.
Given that sense substitution occurs in these changes, and that the different
senses of suffix -v are particular to different contexts, these inferences are not in the
merger class of implicatures discussed above. These data demonstrate the importance
of metonymic inferences at the initiation of semantic change in certain areas, along
with merger inferences, and in contrast to metaphorical inferences.
The general pattern of emerging metaphor may be present in other well-known
changes attributable to inference. For example, the periphrastic future be going to may
set up prediction by implicature. In They are going to sell their wares at the market,
the motion sense They are on their way to market to sell their wares implicates the
future sense They will sell their wares in the market. When the expression generalizes
to contexts in which literal motion towards a destination is excluded, as in Its going to
rain, a metaphorical transfer of MOTION IN SPACE to MOTION IN TIME emerges.

. It could also be argued that some of the individual steps here are metaphorical, although that
is not Heine et al. s intention. To call a hut a house-child, for example, might be said to attribute
metaphorical childhood to the hut in terms of one property which is the grounds for the metaphor,
in this case the property small. Any example of a change which focusses on one element of a lexical
concept may be described in the same terms, where the selected sense component is the grounds
of an analogy between the old and new referent.

137

138 Kate Kearns

The deontic to epistemic shifts in the English modals may show the same pattern. As we have seen, deontic should may implicate epistemic should in all temporal contexts and the same implicature is available for must in some environments.
Epistemic senses also developed for may and can. Bybee et al. (1994:197199) argue
that epistemic possibility may arose by implicature in the Middle English period out of
general ability may, originally physical ability may. (Deontic may was already present
in the Old English period, according to the OED.) Similarly, can developed mental
ability, know-how > general ability > epistemic possibility. The deontic use of can
(expressing permission), not mentioned in the OED, is more limited, and is still considered to be non-standard in some varieties.
The paired deontic and epistemic senses for must, should, may and can emerged
by different pathways, plausibly by implicature in at least some instances. Once established, the whole paradigm unquestionably has the structure typical of metaphor, as
demonstrated by Sweetser (1990), who adapts Talmys (1988) force dynamics model
to analyse deontic and epistemic modals, in addition to other expressions of forcing,
letting and allowing. One might speculate that the late addition of deontic can is analogic to may, triggered by the deontic:epistemic parallelism. The deontic:epistemic
metaphorical mapping is consistent with Lakoff & Johnsons (1980) theory of metaphor
as commonly based on a richly structured concept associated with a field of expressions, rather than with a single word, as in their STATUS = VERTICAL, expressed in
social climber, high/low rank, she married beneath her, upwardly mobile, and so on. The
transfer of inner structure from one concept to another is persuasive evidence for the
presence of metaphor.
Metaphor has long been recognized as a major mechanism for semantic change,
particularly in those semantic domains where according to Lakoff & Johnsons thesis, our very thinking is metaphorically structured, including, for example, domains
such as space, time and motion underlying the meaning of grammaticalized forms.
The basic status of such conceptual metaphors is supported by their appearance in
many languages. Heine et al. (1991) show that for at least some changes in these areas,
metaphor operates only on a larger timescale, emerging indirectly from a series of
implicatural steps.
This contrasts with the fairly direct creative metaphorical semantic changes found
in lexical domains which are by no means conceptually basic, such as the changes in
20th century English studied by Warren (1992). Her examples include in the ball park>
approximate area of an estimate, baseload minimum amount of electric power that a
plant must produce > minimum amount of goods to stay in business, blind-side> to
deal an unexpected blow, footprint > area in which debris from spacecraft may scatter,
and many others.
These two kinds of metaphor present an interesting contrast. On the one hand,
apparently conceptually basic metaphors, sometimes shared across languages, are

Implicature and language change

observed to emerge only indirectly by incremental steps. On the other hand, conceptually elaborate and idiosyncratic metaphors can arise directly. The tension between
these two patterns may reward further investigation.

6. Summary
In summary, if implicature is construed broadly to subsume all the inferential processes
available in language use, then most, perhaps all of the major types of semantic change
can be attributed to implicature. Differentiating more finely, metaphor and metonymy
(which are themselves not always readily distinguishable) produce the ambiguity or
polysemy pattern of added meaning, in contrast to the merger pattern, which is commonly attributed to information-strengthening generalized conversational implicature,
after Horn and Atlas & Levinson. On an alternative view, implicature (narrowly construed) and metonymy are grouped together in contrast to metaphor, on the grounds
that the two former mechanisms are based on sense contiguity, while metaphor is
based on sense comparison or analogy. I have suggested that the relation between
metonymy and narrow implicature is really analogic, and does not support identifying the two mechanisms. Classic Gricean particularized conversational implicature, of
great importance in pragmatics generally, appears to play no part in semantic change,
presumably because it fails the primary requirement of associating new meaning with
a given form.

References
Aijmer, K. (1985). The semantic development of will. In J. Fisiak (ed.): 1121.
Atlas, J.D. & S.C. Levinson (1981). It-clefts, informativeness, and logical form: radical pragmatics
(Revised Standard Version). In P. Cole (ed.) Radical pragmatics: 161. Academic Press.
Bybee, J. & W. Pagliuca (1987). The evolution of future meaning. In Papers from the 7th International
Conference on Historical Linguistics: 109122. John Benjamins.
Bybee, J., R. Perkins & W. Pagliuca (1994). The evolution of grammar: tense, aspect and modality in the
languages of the world. University of Chicago Press.
Fisiak, J. (ed.) (1986). Historical semantics: Historical word-formation. Mouton.
Geeraerts, D. (1997). Diachronic prototype semantics: a contribution to historical lexicology. Clarendon
Press.
Geis, M. & A. Zwicky (1971). On invited inferences. Linguistic Inquiry 2: 561566.
Grice, H.P. (1975). Logic and conversation. In P. Cole & J.L. Morgan (eds.) Speech Acts: 4158.
Academic Press.
Heine, B. (1997). Cognitive foundations of grammar. Oxford University Press.
Heine, B., U. Claudi & F. Hnnemeyer (1991). Grammaticalization: a conceptual framework. University
of Chicago Press.

139

140 Kate Kearns

Hopper, P.J. & E. Traugott (1993). Grammaticalization. Cambridge University Press.
Horn, L.R. (1984). Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature.
In D. Schiffrin (ed.) Meaning, form and use in context: linguistic applications: 1142. Georgetown
University Press.
(1989). A natural history of negation. Chicago University Press.
Kempson, R. (1980). Ambiguity and word meaning. In S. Greenbaum, G. Leech & J. Svartvik (eds.)
Studies in English linguistics: 716. Longman.
Knig, E. (1985). Where do concessives come from? On the development of concessive connectives.
In J. Fisiak (ed.): 263282.
Lakoff, G. & M. Johnson (1980). Metaphors we live by. University of Chicago Press.
Levinson, S.C. (1983). Pragmatics. Cambridge University Press.
Lipka, L. (1985). Inferential features in historical semantics. In J. Fisiak (ed.): 339354.
Sperber, D. & D. Wilson (1986). Relevance: Communication and cognition. Harvard University Press.
Stern, G. (1931). Meaning and change of meaning. Elanders Boktryckeri Aktiebolag.
Sweetser, E.E. (1990). From etymology to pragmatics: metaphorical and cultural aspects of semantic
structure. Cambridge University Press.
Talmy, L. (1988). Force dynamics in language and cognition. Cognitive Science 12: 49100.
Taylor, J. (1989). Linguistic categorization: Prototypes in linguistic theory. Clarendon Press.
Traugott, E.C. (1989). On the rise of epistemic meanings in English: an example of subjectification in
semantic change. Language 65: 3155.
Traugott, E.C. & E. Knig (1991). The semantics-pragmatics of grammaticalization revisited. In
E.C.Traugott & B. Heine (eds.) Approaches to grammaticalization vol 1: 189218. John Benjamins.
Warren, B. (1992). Sense developments: a contrastive study of the development of slang senses and novel
standard senses in English. Almqvist & Wiksell International.

Interlanguage pragmatics
Gabriele Kasper

1. Definition and scope

Interlanguage pragmatics (ILP) represents an intersection of pragmatics and the study
of second language acquisition. Interlanguage, a central construct in second lan
guage acquisition research, refers to second language learners developing (partly
instable, transient) knowledge of a target language (L2) (Selinker 1972). Interlanguage
includes features from L2, from the learners first language or other languages she may
know (L1/Ln), and autonomous features which can be found neither in L1 nor in L2.
Interlanguage theory was first formulated with a view to learners formal linguistic
knowledge. When the first ILP studies appeared in 1979, interlanguage syntax, morpho
logy, and phonology were already well-established areas of investigation.
On an action-theoretical view of pragmatics, ILP has been defined as the study of
nonnative speakers comprehension, production, and acquisition of linguistic action
in L2. While ILP has borrowed most of its research topics, theories, and methods of
data collection and analysis from empirical pragmatics and especially crosscultural
pragmatics, most studies to date have followed the comparative methodology recom
mended by Selinker for interlanguage research, i.e. comparing parallel sets of inter
language, L1, and L2 data.
Unlike other interlanguage studies, ILP studies have predominantly focused on
second language use rather than development. Like other interlanguage studies, ILP
research has explored English more than any other target language.
2. Pragmatic comprehension
Studies of pragmatic comprehension address pragmalinguistic and sociopragmatic
issues. Pragmalinguistic studies examine how learners understand nonliteral utter
ances, both in terms of propositional content and illocutionary force assignment, and
how they assess the politeness value of conventions of means and form. Sociopragmatic
studies investigate how learners perceive variables in the social context and in the
linguistic act itself.
2.1 Comprehension of nonliteral utterances
Nonnative recipients have full access to the cognitive procedures of conversational
implicature and hence comprehension of indirectness, but they do not always make

142 Gabriele Kasper

full use of their inferencing ability. One problem is that as readers or conversationalists,
learners often rely more on linguistic than contextual cues in message comprehension.
Both the type of implicature and learners cultural and linguistic background influ
ence the success of learners comprehension activity. A prominent research issue in
studies of illocutionary force attribution by native recipients is whether the indirectly
conveyed force is recovered immediately and without the literal utterance meaning,
concurrently with the literal meaning, or subsequent to the literal meaning. Advanced
learners adopted the same comprehension process as native recipients (Takahashi &
Roitblat 1994). However, it remains a question for future research whether less profi
cient learners will take a different route (see Takahashi 1990, for review of studies on
nonliteral utterance comprehension).
2.2 Assessment of politeness
Nonnative recipients distinguish different degrees of politeness in conventions of
means and form, but not always in the same way as native recipients. For instance,
Japanese learners of English perceived modification of requests through modals, tense,
negation, and supportive moves differently from English native speakers. American
but not Japanese judges rated positive politeness strategies in English as more polite
than negative politeness strategies. Learners assessments of politeness values may
be influenced by a number of factors, for instance, transfer from their first language,
or social characteristics of the message receiver, such as age and gender. Nonnative
speakers convergence to L2 sociopragmatic norms tends to increase with the time
spent in the target community: learners of Hebrew, who initially based their polite
ness assessments of requests and apologies on L1, became more tolerant of directness
and positive politeness the longer they lived in Israel (Olshtain & Blum-Kulka 1985).
Quality and quantity of exposure to the target language can account for the differential
politeness ratings given by foreign as opposed to second language learners (i.e. class
room learners of a language which is not a regular means of communication in their
community versus residents of the target community). For instance, Japanese learners
of English as a second language (in the US) gave more native-like ratings to different
request strategies than learners of English as a foreign language (in Japan).
2.3 Sociopragmatic assessment
Nonnative speakers assessment of whether or not a linguistic act is appropriate in
a particular social context may differ according to linguistic and cultural environ
ment. Japanese female learners of English felt that refusing a request or offer was
less acceptable in English than in Japanese. Native speakers of English expressed a
greater need to apologize for an offense when they spoke English than Hebrew, while

Interlanguage pragmatics

Russian learners of Hebrew felt that it was the type of offense rather than language that
determined whether apology was called for. German learners of English rated the
severity of offenses warranting apology as higher than native speakers of British English,
whereas Thai learners differed from native speakers of American English most strongly
on the obligation to apologize. In addition to factors pertaining to the speech act in
question, learners may also differ from native speakers in their assessment of partici
pant factors. For instance, compared to American raters, Japanese learners of English
demonstrated more status differentiation but less differentiation of social distance.

3. Production of linguistic action

Most data-based ILP studies examine how learners perform a particular speech act
in L2. Speech acts studied thus far include, in declining number of studies, requests,
apologies, refusals, compliments, suggestions, expressions of gratitude, invitations,
rejections, expressions of disagreement, corrections, and complaints. Adult learners
have an implicit knowledge of the strategies by which such speech acts can be realized,
but they do not always make full use of this knowledge. Intervening factors may be
their limited linguistic proficiency, negative pragmatic transfer from L1, stereotypical
assumptions about what is pragmatically appropriate in target language contexts, or
cultural resistance, i.e. deviating more or less deliberately from L2 norms in order to
keep themselves distinct from the target community.
One recurring observation is that learners differ in the directness by which they
realize a linguistic act. According to most studies, learners opt for more direct strate
gies than native speakers in comparable contexts, however, the opposite is noted as
well: nonnative speakers of Hebrew expressed requests and complaints less directly
than native speakers, and Japanese learners of English used more indirect request
strategies than American native speakers. Whether learners prefer more direct or more
indirect strategies than native speakers may be a reflection of the relative preference
for directness/indirectness in L1 and L2, but such contrastive effects can be overridden
by the factors listed above.
In addition to specific strategies, learners may differ from native speakers in
their overall politeness approach. For instance, Japanese and Taiwanese learners of
English and Athabaskans favored a negative politeness style where native speakers
of American English preferred a positive politeness approach; conversely, learners of
English with Venezuelan Spanish as L1 relied on positive politeness in apologizing
where US Americans were negatively polite.
In sum, learners have access to the same range of conventions of means as target
language speakers, but they often make different contextual choices from speech
act sets (Olshtain & Cohen 1983). On the other hand, the conventions of form used

143

144 Gabriele Kasper

by nonnative speakers are typically more restricted than native speakers, and even
if forms such as tense, aspect, modal verbs, and modal particles are part of learners
linguistic competence, they may not fully control the pragmalinguistic functions of this
material. Among the conventions of form which seem particularly difficult to acquire are
routine formulae (Coulmas 1981), i.e. prefabricated patterns used in recurrent situa
tions and for specialized functions. While learning a set of highly frequent routines
is not problematic in itself since such routines are learnt and stored as lexical items,
matching them with their pragmalinguistic functions and sociopragmatic context con
straints is often only partially achieved even by highly proficient nonnative speakers.
A number of studies document a tendency in intermediate to fairly advanced
learners to produce longer, more verbose utterances than native speakers. This inter
language feature has been related to learners either lacking, or not having sufficient
control over, routine formulae (Edmondson & House 1991). However, Thai learners
of English both used appropriate routine formulae in apologizing and overused nonroutinized apology strategies. While there is some indication that verbosity, whether
compensatory or additive, may occur in written data only, at least one study reports
prolixity also in conversational data.
The function and communicative effect of overproduction remains a topic for
debate. While in terms of native speaker norms, verbosity is a violation of the Gricean
maxims of quantity and manner, it is doubtful whether native standards should be
applied to assess nonnative linguistic action. It seems likely that nonnative speakers
have a greater need to make their intentions, motivations, and reasoning explicit,
because they need to establish rather than take for granted lack of common ground
with speakers of the target language or other nonnative speakers.

4. Development of pragmatic competence

Compared to the large number of studies on various aspects of pragmatic performance
by nonnative speakers, the body of literature on pragmatic development is considerably
smaller. Of these, most studies are based on a cross-sectional design, few on longi
tudinal data, and only two theoretical proposals have been offered to date to account
for the acquisition of pragmatic competence by adult nonnative speakers (Kasper &
Schmidt 1995).
4.1 Cross-sectional studies
Cross-sectional studies compare how samples of L2 learners at different levels of
proficiency understand or produce linguistic action. While most studies show that

Interlanguage pragmatics

learners can use the same pragmatic strategies as native speakers regardless of pro
ficiency level, proficiency effects are regularly found for the linguistic forms and the
contextual distribution of meaning and form conventions. A problem of the crosssectional studies is, however, that the most frequently employed data collection tech
niques, written questionnaires and roleplays, are too taxing for beginning L2 learners;
hence studies usually include intermediate and advanced learners only. The absence of
proficiency effects in the use of conventions of means may thus be a design artifact.
4.2 Longitudinal studies
That this is in fact the case is supported by the few longitudinal studies on pragmatic
development available to date. Studies reporting on individuals or groups of learners
from the early stages onwards show that, consistent with naturalistic second language
acquisition generally, learners initially rely on a few prepackaged routines, which are
later analyzed into rules and elements that become available for productive use. For
instance, early conventionally indirect requests may be realized through sentence
frames such as can I have or can you, whereas modifications through tense and modal
adverbs (could you perhaps) appear only later. While analyzed use of linguistic material
in various pragmatic functions is a more advanced achievement, it also takes more
proficient nonnative speakers to chunk linguistic forms into pragmatic routines and
have such chunks available for fluent access (Schmidt 1992).
4.3 Theoretical accounts
Theoretical accounts of pragmatic development, based on cognitive theory and research,
have first been conceptualized as models of second language acquisition at large and
then extended to the acquisition of pragmatic competence. One such proposal is
Schmidts (1993) theory of the role of consciousness in pragmatic development. Distin
guishing between acquiring and using pragmatic information, Schmidt suggests that
performing linguistic action often happens without conscious awareness, since much
pragmatic knowledge is routinized and therefore does not require much attentional
capacity in proficient language users. In order to first acquire pragmatic information,
however, learners need to focus attention on relevant features of forms, their prag
malinguistic functions, and the sociopragmatic constraints under which they occur.
Noticing relevant input features is required in order to make the information available
for further processing and storage. Another theoretical proposal is Bialystoks (1993)
two-dimensional model of language use and proficiency. This model views language
learning as developing on two orthogonal dimensions, the analysis of knowledge and
control of processing. Bialystok contends that childrens primary learning task in prag
matics is to develop analytic representations of pragmalinguistic and sociopragmatic

145

146 Gabriele Kasper

knowledge, whereas adult L2 learners mainly have to acquire processing control over
already existing representations.
The two proposals are compatible in that they address different stages of pragmatic
learning: Schmidt is concerned with the conditions of initial intake, whereas Bialystok
considers how acquired pragmatic information is represented and restructured. Both
proposals await empirical testing in future studies.

5. Pragmatic transfer
5.1 Definition
Pragmatic transfer has been defined as the influence exerted by learners pragmatic
knowledge of languages and cultures other than L2 on their comprehension, produc
tion, and acquisition of L2 pragmatic information (Kasper 1992, also for review of
research). Learners and bilingual speakers linguistic action patterns regularly show
some evidence of transfer, but the conditions of crosscultural and crosslinguistic prag
matic influence are far from being clearly understood. It is well established though
that, just as in second language acquisition and use generally, transfer in interlanguage
pragmatics cannot be entirely predicted or explained by a contrastive analysis of L1
and L2 pragmatics, although such relationships (on the assumption they are equiva
lent to learners cognitive representations) factor in the occurrence or absence of prag
matic transfer.
5.2 Positive transfer
Outcomes of transfer can be distinguished according to their relationship to the target
language. When learners production of a pragmatic feature is the same (structurally,
functionally, distributionally) as a feature used by target language speakers in the same
context, and this feature is paralleled by a comparable element in learners L1, this is
referred to as positive transfer. Empirically, positive transfer is often difficult to dis
tinguish from pragmatic universals, but in some cases the distinction is quite clear.
For instance, in all languages examined so far, requests can be performed directly
(e.g. using an imperative, as in Water the plants!), conventionally indirectly (using
a illocutionary force indicators such as routinized sentence frames, as in Would you
mind watering the plants?), or indirectly (hinting at the intended illocutionary and/or
propositional act, as in These plants look pretty dry). When learner use any of these
strategic options, they are therefore likely to rely on universal pragmatic knowledge
of request realization. However, when learners interlanguage use features particular

Interlanguage pragmatics

pragmalinguistic means which occur in their L1 and L2 but are not universal, this is a
good case for positive transfer. For example, learners reportedly transfer the past tense
of modal verbs from Danish and German to English (as in L1 Danish kunne du laane mig
dine noter, L1 German knntest Du mir Deine Vorlesungsmitschrift leihen, L2 English/IL
could you lend me your notes).
5.3 Negative transfer
Despite the obviously important role of positive pragmatic transfer in L2 learners
linguistic action, ILP research has paid much more attention to the opposite pheno
menon. Negative pragmatic transfer is observable when a pragmatic feature in the
interlanguage is (structurally, functionally, distributionally) the same as in L1 but
different from L2. For instance, Japanese learners of English may express gratitude by
saying Im sorry, a negative transfer from the Japanese routine formula sumimasen.
In Japanese, apologetic expressions are often used in conveying gratitude when the
speakers emphasis is on the incurred debt, especially when the benefactor is the statushigher person. Danish learners frequently transfer the syntactic request strategy modal
verb + interrogative + negation to English, as in cant you clean the kitchen (from
Danish kan du ikke ryde op i koekenet), apparently not realizing that the negation carries
an overtone of exasperation in English whereas in Danish it mitigates the request.
Strategies of speech act realization are particularly prone to negative transfer. For
instance, Japanese learners of English may use statements of principle (Inever yield
to temptations) as refusal strategies, or information questions as rejections, warnings,
or requests for action. German learners of English tend to extract themselves from
conversation by referring to some specific obligation of their own, following the
German convention (I must pick up Franz from kindergarten), while native speakers
of English prefer vague routines that refer to the other persons needs (I mustnt keep
you any longer). Studies of Chinese-English discourse have shown that pragma
linguistic transfer operates not only at the act or turn level but also in the sequential
organization of discourse.
In addition to negative transfer of pragmalinguistic knowledge, the sociopragmatics
of L1 may be transferred to L2 communication. Whether or not learners perform a
particular linguistic act may be based on L1 preference patterns. Japanese learners
of English expressed reluctance about refusal in an American context because they
would comply in Japanese. Nonnative speakers of English with a Chinese language
background tend to reject compliments, based on the normative preference in Chinese.
Sociopragmatic assessments of interlocutor relationships are also prone to crosscultural
transfer. Japanese learners of English modeled their status perceptions in unequal
power relations on Japanese norms. A related phenomena, though sociolinguistic

147

148 Gabriele Kasper

rather than sociopragmatic, is Japanese students reluctance to call their gaijin teachers
by their first names, which is seen as disrespectful to the sensei.
5.4 Transferability
While the phenomenon of pragmatic transfer is well attested, the conditions of trans
fer and especially the interaction of different factors are less clearly understood.
In interlanguage studies of morphosyntax and phonology, structural properties of
L1 and L2, both objective features and their subjective perception by native and non
native speakers, have been identified as predictors of transferability. In ILP, only one
study has been carried out to date with an explicit focus on transferability. Takahashi
(1995) reported that Japanese learners of English found several indirect request strategies
differentially transferable, and that their transferability perceptions interacted with the
degree of imposition implied by the requestive goal. Since transferability perceptions
were often not consistent English native speakers judgments of contextual appropri
ateness, and learners proficiency had little effect on transferability, it seems likely that
theses college students received too little pragmalinguistic and sociopragmatic input
in their English classrooms.
In addition to structural factors, nonstructural factors impact on pragmatic transfer.
Whereas Takahashi found little influence of L2 proficiency on transferability, actual
transfer varies mostly with learners proficiency level. There is evidence that prag
matic transfer may correlate negatively or positively with proficiency, resulting in four
options, each of which has some empirical support: (1) Low proficiency learners may
transfer L1 conventions of means or form to L2 without realizing the pragmalinguistic
meaning of these structures, such as Japanese learners of English using Im sorry to
express gratitude. (2) Low proficiency learners may not transfer an L1 convention of
means or form because such transfer overtaxes their linguistic competence in L2. For
instance, learners of Hebrew did not transfer English indirectness strategies because
their Hebrew interlanguage did not yet include the necessary complex structures.
(3)Advanced learners may transfer conventions of means and forms because their L2
linguistic competence enables them to do so, but they lack the pragmalinguistic and
sociopragmatic knowledge to correctly assess the functional and contextual equiva
lence relations between L1 and L2. Examples are transfers into English of more speakeroriented, explicit, and specific strategies of linguistic action by German learners and
the more ritualistic, vague, and implicit strategies used by Japanese nonnative speakers.
(4) Advanced learners may not transfer L1 conventions and means even though they
have the required linguistic L2 competence because they believe that such transfer
would not be pragmalinguistically or sociopragmatically successful according to L2
pragmatic norms. An example is Japanese learners of English opting for less polite

Interlanguage pragmatics

refusal strategies than in Japanese because they believed that refusing is more socially
acceptable in English and therefore requires less mitigation.
Rather than descriptively ascertaining pragmatic transfer, future research will have
to focus on the structural and nonstructural factors that promote or depress crosslin
guistic and crosscultural influence in learners linguistic action.
6. Communicative effect
It is important to distinguish negative pragmatic transfer from miscommunication or
pragmatic failure. Positive and negative transfer describes the outcomes of putative
cognitive processes, operationalized in terms of the structural, functional, and con
textual relationships between pragmatic features in L1, L2, and interlanguage. Positive
transfer is a converging relationship, negative transfer one where interlanguage and L1
converge and L1 and L2 diverge. While it is true that negative pragmatic transfer can
cause miscommunication, there is no logical or empirical reason that it has to, and
certainly negative pragmatic transfer is not the same as pragmatic failure. Just as the
equation divergence L1L2 = negative transfer does not hold up because, as we have
seen, for a variety of reasons learners may opt against transfer, the analogous equation
negative transfer = pragmatic failure does not bear out. Whether induced by negative
transfer or other causes, divergence in nonnative speakers communicative style or
pragmatic strategies from those of the target community is not in and of itself a source
of miscommunication. On the other hand, there is abundant evidence that divergence
can cause more or less serious pragmatic failure, from small misunderstandings that
can easily be ignored or repaired to major communication breakdowns with disas
trous consequences for participants, especially those in the less powerful position. In
high-stake interactions such as gate-keeping encounters in medical, educational, and
legal settings, insufficient coordination of knowledge and action between participants
may both be the cause and result of pragmatic failure.
Whether or not pragmatic divergence leads to miscommunication, and if it does,
how disruptive the incidence is, depends on a variety of factors. One is whether the
divergent behavior occurs at the pragmalinguistic or sociopragmatic level. It has been
noted that pragmalinguistic divergence (discrepancies in the use of linguistic material
for illocutionary and politeness functions which are differently conventionalized in L2)
is less serious than sociopragmatic divergence (discrepancies in the assessment of con
text factors, especially those relating to power and solidarity) because interlocutors
tend to regard the former as a linguistic problem, while they may well see the latter as
indicative of the speakers poor manners and moral character (Thomas 1983). Neither
divergence needs to be problematic in itself if the social values indexed by the divergent

149

150 Gabriele Kasper

behavior are acceptable or perhaps even appreciated by the recipient. For instance,
the more egalitarian, positive politeness style often practiced by Western teachers in
Japan may be welcomed by some students because it offers them an enjoyable and
productive alternative to the hierarchical participation structure of the traditional
Japanese college classroom. As long as interlocutors experience pragmatic divergence
as complementary rather than conflicting, it may be regarded as just another way of
doing things. While it is true that a shared communicative style facilitates successful
communication, miscommunication among competent adult members of the same
speech community is a regular fact of everyday life (Coupland, Giles& Wiemann 1991)
rather than a particular hazard of intercultural communication. Intercultural commu
nication is more vulnerable to pragmatic failure than intracultural communication
only when relevant knowledge and strategies of linguistic action are not shared; this
however cannot generally be assumed, as successful interpersonal and professional
interaction between members of different speech communities amply demonstrates.
Recent studies have therefore emphasized that context, identity and intersubjec
tivity are negotiated by participants in situ, and that conversational process and outomes
are more appropriately seen as locally emerging, constructive enterprises than just as
reflecting participants intracultural frames of reference. While participants partly rely
on L1 based strategies and interpretive frameworks, intracultural pragmatic knowledge
is modified or suspended in order to meet the relational and transactional demands
of intercultural encounters (see Janney & Arndt 1992, for a theoretical proposal and
Piirainen-Marsh 1995, for review and a key study).

7. Pragmatic norms
The effect of linguistic action on participants is closely related to the issue of pragmatic
norms in research and teaching. Since the dominant research model in interlanguage
studies is comparative, a yardstick is needed against which to measure L2 learners
pragmatic knowledge and behavior. Likewise, goals for second or foreign language
teaching have to be specified. In both cases, the more or less implicit assumption is
that native speakers of the target language are the relevant population to serve as a
model for nonnative speakers. In ILP research, this is particularly obvious because, as
noted above, most studies compare learners linguistic action patterns against those of
L2 native speakers.
The native speaker norm in ILP has been problematized from different angles.
1. Determining such a norm is difficult because of the sociolinguistic variability in
native speaker behavior. While selecting the variety or varieties most relevant for
a particular learner population in a principled manner is not straightforward for

2.

3.

4.

5.

6.

7.

Interlanguage pragmatics

any target language, it is a particularly daunting task with respect to English, the
language most studied and taught worldwide.
It would be unrealistic to posit an ideal communicatively competent native speaker
as a target for L2 learners since communication amongst native speakers is often
partial, ambiguous, and fraught with potential misunderstanding.
Little is known about adult L2 learners ability to attain native proficiency in prag
matics. While it seems unlikely that a parallel construct to an innate language
acquisition device will be proposed for pragmatic competence, it is still possible
that early and sustained contact with the target language and culture is required
in order to attain native pragmatic knowledge and skill, as is suggested by the
many studies documenting the nonnative pragmatic behavior of advanced learners.
Ifthis was the case, it would be futile to posit a native speaker norm for language
teaching to adult L2 learners, although such a norm would be meaningful for
ILP research.
Learners may not aspire to L2 native speaker pragmatics as their target. Foreign
language learners may not feel the effort is worth their while, since they do not
intend to become part of the L2 community; second language learners such as
immigrants may opt for partial divergence from the pragmatic norms of the target
community as a strategy of identity maintenance.
L2 native speakers may perceive nonnative speakers total convergence as intrusive
and inconsistent with their role as outsiders. Some measure of divergence may be
appreciated as a token of nonmembership.
The communicative style developed by nonnative speakers in interaction with L2
native speakers or other nonnative speakers may significantly differ from that of
L2 native speakers.
Since L2 learners are bilingual or multilingual speakers by definition, the only
reasonable norm for them is a bilingual/multilingual rather than a monolingual
L2 norm.

8. Language teaching
There is not much research examining the effect of language instruction on pragmatic
development, but the few studies to date are encouraging. Adult learners respond well
to a combination of consciousness raising activities in connection with pragmatic
comprehension and production, metapragmatic information about sociopragmatic and
pragmalinguistic target language norms, and communicative practice (Wildner-Bassett
1984, for a key study). In fact, it seems doubtful whether children or adults can acquire
pragmatic competence without some direct teaching, and assistance in noticing relevant
information in the input. The literature on language socialization and developmental

151

152

Gabriele Kasper

pragmatics suggests that children are taught about politeness, form-function mappings
and contextual appropriateness of conventions of means and form (DuFon 1994, for
review). Adult L2 learners in a second language environment are in a better position
to acquire relevant pragmatic information than foreign language learners because they
have access to richer and more frequent input; however, depending on L1 background,
teaching can be facilitative or even necessary for learners to notice pragmatic informa
tion (Bouton 1994). Foreign language learners are at a double disadvantage: not only
do they have less exposure to the target language, but since the action patterns and
discourse structure of traditional classroom interaction are quite different from those
of ordinary conversation, language classrooms provide little opportunity for learners
to obtain relevant input and to produce the variety of linguistic action and politeness
functions required for communication outside the classroom. However, these draw
backs can be compensated to some extent through pragmatically focused curricula,
student-centered classroom activities, and teaching materials providing pragmatic
information (House 1995). The potential of audiovisual and electronic media for prag
matic consciousness raising and communicative activities is particularly rich and still
awaits full exploration.

9. Research methods
Because most studies are based on a comparative design, the predominant data type in
ILP is some form of elicited data (Kasper & Dahl 1991, for review). For comprehension,
the most frequently used instrument is questionnaires including multiple choice, rank
ing or rating tasks. Production of linguistic action has mainly been assessed by means
of discourse completion questionnaires, fewer studies have employed some form of
simulation, or open roleplay. Multimethod approaches are common, often combining
sociopragmatic assessment and pragmalinguistic comprehension or production data.
For instance, a sequential research approach for a study on speech act realization by
native and nonnative speakers will include the following phases: (1) observation of
authentic interaction; (2) choice of contexts; (3) assessment of sociopragmatic factors;
(4)elicitation of production data through role plays or discourse completion question
naires; (5)informants verbal reports on the production task.
There are fewer studies based on authentic discourse. These studies are typically not
or only partially comparative, and mostly report on interaction in institutional settings.
Since the joint management of transactional and interpersonal goals in intercultural
encounters is first beginning to be examined, future research will have to provide
microanalyses of such encounters in a variety of institutional and noninstitutional set
tings, especially those conducted in a language other than English (for a comparative
evaluation of elicited and authentic data in ILP, see Bardovi-Harlig & Hartford 1993).

Interlanguage pragmatics

10. Further reading

A recent collection of articles on ILP is edited by Kasper & Blum-Kulka (1993). Some
important early studies appeared in Larsen-Freeman (1980) and Wolfson & Judd (1983).
The first monograph is Kasper (1981), a recent book-length treatment Trosborg (1995).
Studies in Second Language Acquisition 18:2 is a Special Issue on the development of
pragmatic competence.

References
Bardovi-Harlig, K. & B. Hartford (1993). Comparing observational and elicited data in inter
language pragmatics. Paper, Research methods in interlanguage pragmatics, 10th AILA Congress,
Amsterdam.
Bialystok, E. (1993). Symbolic representation and attentional control in pragmatic competence. In
G.Kasper & S. Blum-Kulka (eds.): 4359.
Bouton, L.F. (1994). Conversational implicature in the second language. Journal of Pragmatics
22,157167.
Coulmas, F. (ed.) (1981). Conversational routine. Mouton.
Coupland, N., H. Giles & J.M. Wiemann (eds.) (1991). Miscommunication and problematic talk. Sage.
Dufon, M.N. (1994). The role of input in the acqusition of politeness. University of Hawaii Working
Papers in ESL 13.
Edmondson, W.J. & J. House (1991). Do learners talk too much? In R. Phillipson, E. Kellerman,
L.Selinker, M. Sharwood Smith & M. Swain (eds.) Foreign/second language pedagogy research:
273286. Multilingual Matters.
House, J. (1995). Developing pragmatic fluency in English as a foreign language. Studies in Second
Language Acquisition.
Janney, R.W. & H. Arndt (1992). Intracultural tact versus intercultural tact. In R.J. Watts, S. Ide &
K.Ehlich (eds.) Politeness in language, 2141. Mouton de Gruyter.
Kasper, G. (1981). Pragmatische Aspekte in der Interimsprache. Narr.
(1992). Pragmatic transfer. Second Language Research 8: 203231.
Kasper, G. & S. Blum-Kulka (eds.) (1993). Interlanguage pragmatics. Oxford University Press.
Kasper, G. & M. Dahl (1991). Research methods in interlanguage pragmatics. Studies in Second
Language Acquisition 13: 215247.
Kasper, G. & R. Schmidt (1995). Developmental issues in interlanguage pragmatics. Studies in Second
Language Acquisition.
Larsen-Freeman, D. (ed.) (1980). Discourse analysis and second language acquisition. Newbury House.
Olshtain, E. & S. Blum-Kulka (1985). Degree of approximation: Nonnative reactions to native speech
act behavior. In S. Gass & C. Madden (eds.) Input in second language acquisition: 303325.
Newbury House.
Olshtain, E. & A. Cohen (1983). Apology. In N. Wolfson & E. Judd (eds.): 1835.
Piirainen-Marsh, A. (1995). Face in second language conversation. Studia Philologica Jyvaeskylaeensia 37.
Schmidt, R. (1992). Psychological mechanisms underlying second language fluency. Studies in Second
Language Acquisition 14: 357385.
(1993). Consciousness, learning and interlanguage pragmatics. In G. Kasper & S. Blum-Kulka
(eds.): 2142.

153

154 Gabriele Kasper

Selinker, L. (1972). Interlanguage. IRAL 10: 209230.
Takahashi, S. (1990). Exploring the comprehension process of nonliteral utterances and some impli
cations for automaticity. University of Hawaii Working Papers in ESL 9: 6797.
(1995). Pragmatic transferability. Studies in Second Language Acquisition.
Takahashi, S. & H. Roitblat (1994). Comprehension of nonliteral utterances by nonnative speakers.
Discourse Processes.
Thomas, J. (1983). Cross-cultural pragmatic failure. Applied Linguistics 4: 91112.
Trosborg, A. (1995). Interlanguage pragmatics. Mouton de Gruyter.
Wildner-Bassett, M. (1984). Improving pragmatic aspects of learners interlanguage. Narr.
Wolfson, N. & E. Judd (eds.) (1983). Sociolinguistics and language acquisition. Newbury Park.

Jargon
Luisa Martn Rojo

1. Introduction
Jargon is an ill-understood linguistic phenomenon, despite a number of sociolinguistic
studies and despite the use of the term jargon in common parlance to refer to a wide
variety of substandard, deviant and closed forms of language usage. In this contribution, a clarification of the phenomenon will be attempted by means of a discussion of a paradigmatic case: delinquent jargon (largot du milieu; see Martn Rojo
1993, 1994). This discussion will extend over Sections 2, 3, and 4. In Section5, I will
consider some other phenomena which can be included under this label: varieties
of jargon jargons of professions, as well as other closely-related linguistic varieties
in particular, juvenile slang, and the argot gneralis common slang (Franois 1968,
1990). The panorama presented in the first part focuses especially on the study and
explanation of the dichotomies normal/abnormal, correct/incorrect; it will contribute to the understanding of all these varieties and to some insight into the sociolinguistic order.
Delinquent jargon is understood here as an extreme example of linguistic variation (Halliday 1978). By taking such an extreme case of a social dialect, we are able
to see more clearly related phenomena, thereby gaining a better understanding of the
general concept of linguistic varieties.
It also seems appropriate historically to begin here. The term jargon originally
referred to an unintelligible type of speech, and was mainly used to designate the slang
of members of the underworld, only later being applied to other linguistic varieties,
such as professional jargons. Its evolution was parallel and coincidental in different
countries. By the sixteenth century, its meaning seems to have been completely established in Italian (gergo or zergo), Spanish (jeringona or jerga), French (jargon), English
(gibberish or gabbling), and Portuguese (geringona).1

. At that time, many synonyms appeared in different countries: in Spain, during the Golden
Age (the sixteenth and seventeenth centuries), the term germana (group of thieves and beggars)
takes this meaning; later on, it was replaced by the terms cal jergal and cal delincuente (stressing
the role of the language of Spanish gypsies as a source of inspiration); calao in Portuguese (also
emphasizing the gypsies contribution); argot in French (following the same process as germana);
furbesco in Italian; cant in English; Rotwelsch in German; and so on.

156 Luisa Martn Rojo

However, there are other reasons why delinquent jargon may be a paradigmatic
and illuminating example of a number of related linguistic phenomena. These reasons
have to do with the capacity of discourse (e.g. scholarly discourse) to create and position
concepts and subjects.
Power relationships entail the building up of fields of knowledge while, simultaneously, knowledge involves and constitutes power relationships (Foucault 1977).
As we will see, a connection could be established between the production of some
discourses about jargon and social changes, and between these discourses and changes
in social practice, among them the appearance and consolidation of new models of
punishment. Given the deviated and subversive values associated with this marginal
variety, most of the traditional theories and explanations seem to have contributed to
justifying and legitimizing different procedures of social exclusion and stigmatization.
This was achieved in a number of ways: by transferring to a linguistic domain the fear
of social disorder and anomie, and the ideal of homogeneity; by reinforcing the association between delinquency and abnormality; and by maintaining and contributing
to the fallacy which identifies delinquency and lower social classes. An investigation
of this transfer in relation to the phenomenon of jargon may lead to explanations for
similar negative reactions produced against other social varieties of language.
As I intend to show in this paper, it is difficult to explain how, after the development of the field of sociolinguistics and, especially, after the insightful approach to
jargon in the work of Michael Halliday (1978), many of the traditional and discrediting
ideas about jargon are still pervasive. Even though we already have the conceptual
apparatus required, and even though current studies refute the traditional approach to
delinquent jargon (Culbertson & Fortune 1986; Martn Rojo 1989; Klerk 1990; Michael
1993; Burke 1995; Beier 1995; Sanmartn 1996), prejudice still remains extremely deeprooted. As I would like to show, understanding the reasons for this permanence could
be relevant for a more complete and unbiased view of linguistic variation.
2. S
 tudies of delinquent jargon and the introduction
of a value-giving measure
In many countries in Europe, from the fifteenth century to the present day, we find
evidence of a controversial interest in jargon. Franois Villon wrote his ballads in the
argot des Coquillards. Later, the jeringonza became widespread in Spanish Literature
of the Golden Age (sixteenth and seventeenth centuries).2 A similar phenomenon
. This tendency gave rise to some specific genres closely related to jargon, like the Novela picaresca
(especially Guzmn de Alfarache by Mateo Alemn) and the poems known as jcaras. Among the
authors who used jargon in their plays, Cervantes (in Rinconete y Cortadillo) and Quevedo (in his
jcaras) stand out.

Jargon

appeared in other countries, like Italy, England, Germany and Poland. From these centuries, we have inherited the first dictionaries of jargon.3
Interest in the slang of the underworld grew in the nineteenth century (especially
in relation to the language of gypsies), as a part of the interest in nationalism and
cultural and linguistic differences. However, apart from Vidocq (1837) in France and
Av-Lallemant (18581862) in Germany, complete studies of jargon did not appear
until the end of the century. At that time, it became evident that the process of objectivation (Foucault 1977), by which delinquents were considered a natural but deviant
class, must have started in the second half of the eighteenth century and been completed by the end of the nineteenth. In the works of Salillas (1896), Niceforo (1897,
1912), Sainan (1907), Dauzat (1917), Wagner (1924), and Ghnter (1919), we already
find not only examples of jargon or lists of vocabulary and linguistic descriptions, but
also a conception of this sort of speech, which extended the prevalent view of delinquents themselves to their language.
The simultaneous development of these studies and the creation and establishing
of criminology and other sciences of man, such as clinical medicine, psychology and
psychiatry, cannot be a mere coincidence, and are explained by Foucault (1977) in
terms of a shift in the understanding of illegal practice due to the socio-political and
economic shifts which took place in the second half of the eighteenth century. These
transformations had different consequences among them, the higher juridical and
moral value placed on property relations and a correlative extension and refinement
of punitive practices (stricter methods of surveillance, a tighter partition of population, more efficient techniques of locating and obtaining information). The extension
and refinement of punitive practices and a new organization of the power to punish
were the basis of the appearance of new fields of knowledge, especially criminology.
In the second part of the eighteenth century, then, we witness a double process: on the
one hand, an epistemological thaw through a refinement of power relations; on the
other, a multiplication of the effects of power through the formation and accumulation of new forms of knowledge (Foucault 1977:225).
From this, two processes of objectivation emerged at the end of the eighteenth
century. First, the codification, definition of offences, the fixing of a scale of penalties,
rules of procedure, definition of the role of magistrates (Foucault 1977:102); to carry
out all this, new discourses were generated (exhaustive codes, sufficiently precise for
each type of offence, which entail the consolidation of a legal jargon).

. E.g. the Nuovo modo de intendere la lingua zerga, New way of understanding jargon; 1545; in
Dutch, Der Fielten Vocabulaer, The rogues vocabulary; 1563, in English, Thomas Harmans Caveat
for Common Cursitors, 1566 and Heads Canting Academy, 1673; in Spain, the vocabulary in Juan
Hidalgos Bocabulario de Germana, 1609 (written, in fact, by Cristobal de Chaves); in France,
Le jargon de largot reform, 1628.

157

158 Luisa Martn Rojo

Second, there was a scientific objectivation, by which the criminal was defined as
an object of study. This second process also required the production of new discourses,
in which a space of differentiation was established between normality and abnormality
and in which the rules to be followed were made explicit. These rules sought to homo
genize and to classify all kinds of behaviour, both linguistic and non-linguistic.
Classical studies on jargon can be seen as simultaneously the effect and agents
in this process. On the one hand, they reflected this objectivation by the recreation
of the homo criminalis; on the other, they contributed toward affirming the deviant
nature of criminals.
In Lombrosos theory (1876), particular physical features which progressively
became associated with illness and malformation were attributed to criminals. In
this way, they began to be assimilated to the madman and the sick person that is,
to a sort of destruction (a dangerous germ) that lives inside another living but, in this
case, healthy organism, the society. The majority of the studies devoted to jargon reproduced the most widespread views on society in the nineteenth century (especially social
Darwinism and evolutionist theories): social order depends on the balance between
a process of differentiation and a process of integration or control. On the basis of this
view, we find a radical distinction between the legitimate forces of order (society, integration, control, normality) and the illegitimate forces of disorder (tension, illegitimate
violence, abnormality). Social order was considered to be a very fragile entity, threatened
by social differences and changes: difference was a permanent cause of disorder, as
changes are always a source of tension.
The conception of jargon which was built up during this period is very close to
this negative view of difference and the identification of its creators with destructive
forces. Delinquent jargon, then, was seen in Niceforos words as an infected lymph,
which could poison and corrode all the organism (Niceforo 1897/1972: 174). This
opinion is still echoed in contemporary folk sociology and folk linguistics.
In the early studies, linguistic features attributed to it were used as evidence for
this image of jargon and its speakers (though the features mentioned are no different
from those found in other jargons and in language in general). Thus, Salillas (1896:33)
criticizes the way authors such as Lombroso (1876:487) focus on linguistic features
which link jargons either to primitive languages (onomatopoeia, archaisms both
understood as atavistic forms) or to the abnormality of their speakers (deformation
of words, crude metaphors). Salillas notes that onomatopoeia appear in jargon no
more frequently than in standard language. As regards archaisms, writers on jargon
paradoxically defend the fact that jargon can be characterized both by their presence
and by an extremely fast evolution, explained by the need for secrecy.
Fast evolution is one of the features of jargon which has been exaggerated. Other
traits which are undoubtedly present in jargon, like the proliferation of synonyms
and the deformation of common forms, are taken as evidence to prove the parasitic
character of jargon vis--vis the standard language, its degrading role, and its chaotic

Jargon

structure (proliferation and overlapping of forms). In this way, jargon is understood as

a deviation, not as a process of variation (a dterritorialisation de la langue majeure, in
Deleuze & Guattaris words (1980:127139)). The main action of this parasitic variety
is the corruption, and debasing of language. (Interestingly, its action is opposed to
legitimate forces of controling language, e.g. by scientific academies).
However, it is its function as a secret language devised purposefully for criminal
ends which has been considered its most distinctive feature (Niceforo: 1912:98).4
This focus on premeditation implies that the primary aim of jargon is secrecy
and deception. And what is more important, the desire for secrecy serves to explain
the appearance of the phenomenon itself, which, in consequence, is understood as an
illegal and destructive tactic (see Franois 1968:625; Halliday 1978:166; and Martn
Rojo 1989, 1991).
From a sociological point of view, the emphasis on secrecy is linked to a particular
view of both the social group which uses jargon and the society as a whole. The presentation of jargon as a defensive strategy shows that the social basis of this linguistic phenomenon is understood as an isolated group, able to deliberately create its lexical units
and modify them if they come to be known outside the group. Such a view requires
the postulate of an almost clandestine organization or secret society, which builds a
linguistic area of exclusion and inclusion in order to unite and protect its members.5
This picture of a well-organized brotherhood has been maintained to the present
day. It was nourished in the early nineteenth century by the ideas of social Darwinism:
the underworld seen as another people within the same people, with its own habits,
instincts, morals, as Marquet-Wasselot remarks in 1832 in La Ville du refuge (quoted
by Foucault 197:253). However, as Foucault remarks, now that society was no longer
divided into estates (as in the Ancien Rgime), and in the context of the changes in
the power to punish, this view permits a kind of zoology of social sub-species to
develop and contributes to the creation of a new object, analyzed as a morbid syndrome: the delinquent.
These classical studies of jargon had normalizing effects and they produced many
of the ideas, still prevalent today, about these linguistic varieties. First, they contributed to the construction of a new social actor: the delinquent. At the same time, they
established a pernicious association between crime and social class and between the
breaking of the law by the lower social class and abnormality. The studies reason

. Lintention de demeurer secret afin de protger le group argotier, ou lintention de natre dans
lombre -la premeditation- forme la marque didentit de largot (Niceforo: 1912:98).
. This image is supported in part by literature in fact, some plays present the underworld as
a royaume (kingdom) or cofrada (fraternity), which is well organized and hierarchically structured (thus, in the Jargon de largot reform (1628), which also presents such a view, the archisuppts,
archi-mates or archi-complices, create and expand their jargon).

159

160 Luisa Martn Rojo

that as this destructive sub-species in which irrationality, violence, and deviation are
embodied is located within society, measures of control have to be increased to defend
the community.
At the linguistic level, two normalizing effects stand out. First, there is the result
of a negative view of linguistic differentiation, which introduces a value-giving measure and draws the line between normal and abnormal linguistic forms; between normal and abnormal social dialects; and between legitimate (healthy) and illegitimate
(destructive) linguistic tendencies (homogenization vs. heterogeneity). In short, the
traditional approach to jargon normalizes. And second, it allows a process of appropriation of discourse: only legitimate forces and legitimate social groups have a right
to a language. Minorities like gypsies, and socio-professional and dissenting groups
like delinquents, do not; what they have instead is a dangerous and evil jargon. The
importance conferred on premeditation presents the development of jargon as a
strategy and never sees it as a spontaneous form of speech, as a language, as standard
for the group which uses it, in the same way as the norm is for a society as a whole.6 In
general, it was not seen as a discourse of resistance and protest against domination nor
as a discourse against mechanisms of discipline. Jargon is not understood as a territorialisation, i.e. creating and expressing their own rules and values, but as a dterritorialisation, i.e. taking over the territory of the other, thus putting in danger legitimate
language (Deleuze & Guattari 1980).7
3. Interpretations of antilanguages and their functions
3.1 Halliday: The concept of antilanguage
The concept of antilanguage, a term coined by Halliday (1978:165182), entails a
rupture with the theories we have examined.8 Halliday considers jargon as a discourse

. Paradoxically, this is the definition of jargon we find in the sixteenth century Spanish: with
words suited to the understanding and the life of these people (Chaves 1592/1983).
. This is not an isolated example of appropriation of discourse, as Torrione (1994:95112) shows.
Spanish Roman was considered not to be a language and was also accused of being a premeditated
strategy for deceiving good people. Palmireno (1573:5657) refers how as gypsies do not understand Egyptian, nor Greek, their language was just a trick, to prevent outsiders from understanding
them. As a consequence, the use of Cal (Spanish Roman) was forbidden, and jargon excluded
from the universe of legitimate discourse; a discredited discourse can be easily silenced (for a
similar process affecting slang and dialectal varieties, see Bourdieu 1991; and Section4).
. Besides the new understanding found in Hallidays approach, which I consider especially
important, there are many other studies of antilanguages found in different countries and at different

Jargon

of resistance, a linguistic variety generated by antisociety. Antisociety itself is seen as a

society that is set up within another society as a conscious alternative to it. Antisociety
is, then, a mode of resistance, which takes the form either of passive symbiosis or of
active hostility and even destruction (1978:164). From Hallidays point of view, an
antilanguage stands to an antisociety in the same relation as does a standard language
to conventional society.
This means that jargon should be considered in the same way as any other instance
of the prevailing sociolinguistic order. Furthermore, jargon presents us with an object
of study which, instead of being peripheral, can be seen as a most clarifying paradigm
for the investigation of varieties. In this way, there is continuity between language and
antilanguage, as there is continuity between society and antisociety. The most striking
linguistic features of this type of jargon, partial relexicalization (creation of new words
for old ones) and overlexicalization (proliferation of synonyms), are normal features
of the change which is taking place constantly in any language, whatever its status
from standard to pidgin. All these creations are metaphorical variants because they
do not have the same meaning as everyday forms. It is this metaphorical character
that defines the antilanguages (Halliday 1978:175).
There are two main sources of the characteristic features of antilanguages. First,
these varieties shift the normal balance between the different functions internal to the
linguistic system, moving from the experiential mode of meaning towards the interpersonal and textual modes.9 In this way, overlexicalization can be understood as the
effect of a greater orientation in these two directions: towards the interpersonal mode,
where knowing and using many special group words shows the importance of membership; and towards the textual mode, foregrounding the message with an abundance
of semi-synonyms which represent different attitudes towards their referents. Both
tendencies are also present to a lesser extend in standard language. Secondly, they
differ from standard languages in the kind of meanings exchanged. As an instance
of a particular sociolinguistic coding orientation, antilanguage can be considered a
systematic pattern of tendencies in the selection of meanings expressed. It is a coding
orientation which confers on the languages of countercultures, as on every social dialect, their reality-generating power.
The function of antilanguage is to create an alternative reality, which is the source
of an alternative identity for its members, in the same way as every social dialect

periods. See, for example, in addition to the references in the text, for the Spanish-speaking world,
Garca Ramos 1990; Altieri 1972; Sastre 1980; Paz 1994.
. Halliday looks for an explanation of overlexicalization consistent with a functional orientation
in language analysis and rejects traditional views, which explain it as a search for originality, for
liviness, and especially, for secrecy.

161

162 Luisa Martn Rojo

is the embodiment of a mildly but distinctly different world view (Halliday 1978:179).
Antilanguages are not, from this view, professional varieties, because their appearance
is not related to a different social activity, but to the particularities of its speakers (social
origin, experiential and cultural differences). However, among social dialects, jargon
is one which is particularly threatening, because it not only conveys, but creates and
maintains a counter-reality, not just a subcultural variant (Halliday 1978:181). This
means that what is relevant, from Hallidays point of view, is not the distance between
the two realities, but the tension between them: a counter-reality is set up in opposition
to an established norm (as in the looking-glass world). This opposition explains the
subversive values of antilanguages, and, as a corollary, the aggressive attitude towards
them, which is commonly manifested by speakers of standard dialects.
Subjective reality can be transformed by language, or, in our case, by antilanguage,
and this involves a process of resocialization of speakers.
3.2 Sociolinguistic functions of delinquent jargon
On the basis of an analysis of delinquent jargon in its practical conversational usage,
four different functions can be distinguished (Martn Rojo 1985, 1989, 1993). To some
extent, all these functions can be discovered in other varieties of jargon:
i. Differentiation and opposition
Jargon, like every social dialect, appears as a process of linguistic differentiation, which
reflects and reinforces a correlative process of social differentiation. In delinquent
jargon, this function is related to the anomic attitude, linguistic or otherwise, of its
speakers towards the law, and to the rejection of the submissiveness which implies
the adoption of legitimate and standard ways of speaking. From a sociological point
of view, such a rupture allows speakers an alternative definition of identity (signalling
that they belong to a distinctive group, which is characterized by a particular way of
speaking), while from a linguistic point of view, it can be considered as a process of
dterritorialisation of the standard language. When individuals assimilate the language
and the forms of the majority, they adopt the linguistic norm (standard language and
standard forms); when individuals reject this social and linguistic homogenization,
they change their language. Both tendencies coexist in individuals and both give rise
to linguistic varieties (langues majeures vs. langues mineures, or jargons) (Deleuze &
Guattari 1980; see also Bourdieu 1991:9496).
A number of linguistic features of jargon seem to be especially related to this tendency towards breaking linguistic norms and the same is true for juvenile slang (for
a more systematic study of the linguistic features of antilanguages and their dynamic
and open correlation with these functions, see Martn Rojo 1993). It is interesting,
for example, that this type of language is not only relexicalized in the areas which are
particularly related to the activities of the criminal way of life (e.g. types of criminal

Jargon

acts, types of criminals, victims, police, penalties, and so on), but it also affects many
other aspects of everyday life (meals, games, sex, family structure, and so on). A large
part of the vocabulary is created by a procedure of morphological transformation
(processes like affixing, affix-shifting, compounding, truncation, etc.), suggesting a
process of linguistic appropiation, by which the original shared form becomes a special,
exclusive form belonging only to the group.
In the same way, we find overlexicalization and the proliferation of synonyms,
especially in the areas of vocabulary which seem to be most important for the life of
the group (e.g. in reference to the police or to traitors). In many cases, these forms
are not really synonyms: the choice of one in preference to another shows the user to
have a better knowledge of jargon and its counterculture. However, in some cases, this
proliferation of forms seems to come merely from a desire for change.
ii. Integration and the creation of a private world
Jargon, like a standard language, appears as a means for the regulation of the life of
the group and in-group interaction. By improving in-group communication, jargon
functions as a force towards cohesion, a way of forming the social unit, in the same
way as standard languages do (according to structuralism). Jargon can be understood
in this way as a type of territorialisation because by creating an alternative vocabulary,
it functions as an instrument for knowing, constructing and maintaining an alternative reality, with its own social actors, hierarchies, rules, ways of life, and values. Asan
instrument of knowledge and communication, jargon makes possible a consensus on
the meaning of the social world, which reinforces the social unit. In this sense, it can
be considered an instrument of social integration (see Bourdieu 1991: 166). Given
that, in this case, differences in world view are foregrounded, this linguistic process of
construction of reality entails a correlative process of deconstruction of a shared and
dominant or conventional reality. This deconstruction is a source of tension, which
explains the defensive attitudes and the defensive strategies it brings about. (See below
for defensive values and uses).
The function of creating a private world is related to the coining of new lexical
items, which designate everything connected with the activity of the group (techniques,
tools, professional actions) but also with everything which is part of the way of life
of its creators (objects, values, social actors). This is why some authors call this function referential (Sanmartn 1996) and consider it linked with the professional activity
of the group and with the need to express distinctions and meanings which are not
relevant outside the group. However, the process of creation of lexis does not stop
with professional activities. The whole social network of its speakers is reflected and
expressed by jargon: family relations and the classification of individuals in relation
to their degree of respect for the norms of the group. This tendency includes outsiders,
victims, as well as enemies: the police and other representatives of law and order.

163

164 Luisa Martn Rojo

In these cases, jargon words and standard words are not synonyms but are clearly
opposed by their attitudinal components. As a consequence, they can be considered as
metaphorical variants (see Halliday 1978:166;175177).
If a language fits the reality of its speakers, the possibilities of mutual understanding
and communication are reinforced, while understanding by outsiders is prevented.
The differences between the two realities and between jargon and standard language,
explain how jargon acts as a defensive barrier. Thus, the defensive and cryptic values,
often stressed by authors, can be understood as a subfunction or a subsidiary function of this privatizing function. This function of making a secret language permits a
process of management of information, of which speakers of jargon seem to be particularly conscious. The identity of jargon speakers, their activity, and their way of life
have to be kept secret. However, it is not only the particular activity of the group but
also the tension produced by cultural, social, and moral divergences, which explain
the tendency to block the flow of information outside the group. Some linguistic
phenomena like the use of nicknames (see Goffman 1963), the differential regulation
of euphemisms and dysphemisms, the use of a repertoire of gestures, and sub-codes
created by the inclusion of syllables, are particularly connected with this function
of secrecy. However, these sub-codes are, in fact, very unusual, and only used infrequently (see del Toro 1975). (Management and control of information is connected
with the stigmatizing function I will study below). Such an attitude and the same
function also appear in every case of professional jargon, though the motives behind
them are not necessarily the same. This is similar to the objective of censoring and
establishing a dividing line between sacred and profane knowledge, in Bourdieus
terms. (See Section4 below.)
iii. Liberation
As a consequence of the tension which exists between the two realities, the two ways
of life and the two world views, and as a result of the need to escape from the negative view that standard language projects on jargon and its speakers, jargon appears as a
vehicle of reality maintenance (Berger & Luckmann 1966, adopted by Halliday 1978).
Jargon allows users to avoid the norm, or to avoid normalization, in the Foucaultian
sense. Thus, jargon prevents the internalization of dominant and discrediting discourse:
if speakers internalize this legitimate discourse, they will see themselves as transgressors, asocial, and even abnormal beings. Thus, jargon acts as a discourse of reconstruction, by which its speakers build and rebuild themselves and their images, against
the destructive and discrediting action of the norm: they often present themselves as
victims of social inequality and classify themselves with respect to their own values.
There is a constant need among stigmatized individuals and deviant groups to escape
the norms, since they exclude and reject them.

Jargon

As Goffman (1961:2092) notices, in prison jargon prevents some of the devastating effects of total institutions, which project a negative image of their inmates as
asocial and dangerous beings. The internalization of this social rejection is part of the
rehabilitation work of such institutions. Jargon, by means of relexicalization, creates
a new system of meanings, which builds up a different system of values and protect
individuals from this and other practices.
Thus, an attitude of liberation from the norm and pursuit of autonomy is at the
root of this third function. Jargon, like other social dialects (e.g. juvenile slang), is not
only a discourse of resistance, but a discourse of opposition. (See Section4.)
iv. Stigmatizing
The embodiment of a different world view, an open display of difference, produces the
aggressive attitudes towards nonstandard speech and to its speakers. Linguistic varieties
become a discrediting and stigmatizing sign which transmits social information (viz.
the deviant behavior of their speakers, their opposition to different norms). But it is
also a sign of the existence of a power which has to be opposed: jargon makes manifest
for other groups and for itself the existence of social differences and a refusal to submit, which would be implied by the adoption of legitimate ways of speaking. However,
as Goffman (1961) claims, signs of stigma have different meanings in different groups
because they refer to the same category, but give it different characteristics. In this
sense, the knowledge of jargon has a double value for its speakers. It is highly valued,
because it guarantees membership of the group and the respect for shared values. At
the same time, it becomes a source of danger when communicating with outsiders,
since it is by means of jargon that the group is known and identified. The established
and legitimized distinctions between normal and abnormal linguistic habits fulfill a
political function: contributing to the social integration and legitimation of the dominant group, while devaluing other ways of speaking and stigmatizing their speakers.

4. Varieties of jargon
For the study of the different varieties of jargon, we have to expand the panorama
presented in the previous section and to evaluate the implications of introducing a
value-giving measure into the linguistic domain. We have seen that every linguistic
variety is understood as a deviation and is forced to fit into a set of categories. However, not every variety is evaluated with the same strictness: the different conceptions
of the world and society found in each variety and the prestige and the power of their
speakers have a decisive role in their evaluation.

165

166 Luisa Martn Rojo

In this sense, Bourdieus concept of the linguistic market (Bourdieu 1991) could be
helpful. As Bourdieu points out, the standard language (as a normalized and legitimate
product) became indispensable in the historical construction of the nation. Such an
impersonal and anonymous linguistic form serves uses which have these same properties (official uses: administrative, legal, and so on), and normalizes the production of
linguistic habitus (Bourdieu 1991: 48). To counterbalance legitimation, we find the
devaluation and discrediting of popular modes of expression, slang, jargon, and so on
(especially through the education system).
The process of legitimation plays a decisive role on the understanding of linguistic varieties. In the first place, because it contributes to conforming and imposing an
image of the linguistic space as a hierarchical universe in which one standard form is
accepted as the legitimate form, the pattern by which to measure all others, the place
of consensus, that which is universally known, and to which every other form must
conform. Any deviation means an inferior and incorrect variety.
A second implication of the process of legitimation is the application of dualistic
taxonomies which structure the social world according to the categories of high and
low (a low form of speech); refined and coarse (coarse language) or rude (rude jokes),
distinguished and vulgar, rare and common, well-mannered and sloppy: in short,
categories of culture and nature (Bourdieu 1991:93). So, the process we have studied
in relation to jargon is reproduced in the whole linguistic universe and becomes stronger
when linguistic divergences are perceived as greater. The linguistic market understood
in this way is as remote from the real linguistic market as the pure market of theoretical
economists is from the real economic market, with its monopolies and oligopolies
(Bourdieu 1991:56). Dichotomies are the product of social evaluation and, in consequence, any linguistic classification, being based on this measure, and even linguistic
descriptions of every variety, really need to be re-examined. What researchers who go
in this universe find is, as Halliday points out, a cline, a continuum, a universe in which
differences are not always or, at least, not clearly, seen and so they cannot be isolated.
The situation makes authors like Fowler (1991:60) claim that a variety can be cued
more by stereotypical details than by systematized patterns of variation, and that some
models of dialect and register can be established on this basis.
Among the different phenomena which have attracted the attention of researchers,
professional jargons rank highest. There is a long tradition of research into early forms
of jargon (see, for example, for French professions: Dauzat 1917; the jargon of sailors:
Parry 19481949; Roeding 17941798; Kluge 1911; soldiers: Mausser 1917; Brophy &
Patridge 1930; doctors: Singy 1986; Dirckx 1983; Porter 1995; butchers: Robert
LArgenton 1990; boxers: Choron-Baix 1990; and so on), as well as into new forms,
where emerging jargons, related to new professions, are beginning to receive attention
(especially urban delivery workers, using vans or motor bicycles, who communicate
with each other by radio).

Jargon

Attitudes towards such varieties, and evaluations of them, differ on the basis of
the social conditions of their production and on the social positions of their speakers.
The power for exclusion inherent in these professional varieties, preventing outsiders
from getting into contact with those in authority, seems to be especially disturbing
and to have stronger and more dangerous social effects (see, for example, Porter 1995;
OBarr 1982). The same effect has been pointed out with respect to academic language
(Burke 1995), political propaganda (Herman 1992) and philosophical jargon (see, for
example, Adorno 1963; Bourdieu 1991:137159).
Professional varieties have been considered as registers, with a use linked to an
attitude of defence of the social unit made up by their speakers, and with the distinctive feature of membership of the profession creating that social group (i.e. the
function of differentiation). However, this should not hide the fact that these varieties
always entail a different view and organization of the social world and the expression
of different values, just as we have seen with respect to delinquent groups or antisocieties
(i.e. the function of integration and the creation of a private world). A second phenomenon which has also been studied frequently is that of juvenile slang, traditionally
connected with students (Marples 1940, 1950; Objartel 1985; Bensimon-Choukroun
1991), and more recently with urban tribes and movements among the young in every
country (see, for example, Rodrguez 1986; Coyaud 1990; Lafage 1990; Louis & Primaz
1990; Merle 1990; Yoffe 1992). These varieties are often considered examples of social
dialects. In this case, it is the attitude of transgression of the norm and the pursuit of
differentiation which has been considered the reason for their appearance (i.e. the
function of differentiation and opposition). However, the function of opposition is
clearly accompanied in this case by that of naming things which are specific to a way
of life (hobbies, relevant distinctions among different groups of young people, and so
on; i.e. the fuction of integration and the creation of a private world). And the same is
true of the desire to escape from the normative set of values which traditionally sees
the young as a force which is out of control (see (iii) liberation; for the study of the
process of construction of identity through the use of different languages and varieties,
see Pujolar 1995).
The inversion of values seems to be constant in these slangs, as is the creation of a
linguistic barrier, which is particularly clear in verlan (which depends on a systematic
procedure of inversion and deformation; for example barjot from jobard street urchin;
see Andreini 1985). The presence of the functions of creating a private world and, to
a lesser extent, of liberation, suggests that these varieties are forms which are already
close to delinquent jargon which, in fact, serves for them as a source of inspiration.
The fact that we can distinguish in this type of jargon the functions established above
may help to explain the emergence of a new phenomenon (like the emergence of a new
slang in French banlieues, which is a mixture of French, Spanish, and Arabic forms,
which, nevertheless, is unintelligible to speakers in these communities).

167

168 Luisa Martn Rojo

A final phenomenon which has attracted the attention of scholars is the common slang (Franois 1990), which is linked to popular and casual uses of language.
Researchers tend to point to creativity and a sort of playful expressiveness as explanations for its appearance. At the same time, however, there seems to be a desire to reject
any formal, standard language imposed by a dominant group, as well as to reveal the
stylization of reality that such a language entails (i.e. the opposition function). This
means that this type of jargon may appear and spread in any country at any time.
Thus the panorama is complex and extremely dynamic. One of the most comprehensive efforts to deal with it is found in the different pieces of research carried out by
the CEPLAFE (Centre dtudes du patrimoine linguistique des argots francophones
et trangers), which organizes, promotes and coordinates numerous studies on all
aspects of this subject (see, for example, Descamps-Hocquet, 1989; Franois 1990; and
the working papers produced) and includes a complete set of materials.10
5. Conclusions
I have tried to show that many of the explanations and evaluations of linguistic varieties
such as jargons cannot be understood outside the framework of power struggles, which
are aimed at making people speak (and see the world) in a particular way. They are, in
fact, attempts to impose a particular social order, with its divisions and hierarchies.
Jargons are linguistic varieties that have proved to be very liable for integration in
descriptive and evaluative dichotomies such as normal/abnormal, standard/deviant.
A fairer and probably more accurate perspective should take into account the various
social functions served by jargon and other linguistic varieties and grant them a less
biased sociolinguistic status. Jargon in that sense exposes some of the deep pretheorical
bases of language studies, and can be put to use as a remedy against them.

References
Adorno, T.W. (1973). The jargon of authenticity. Northwestern University Press.
Altieri Barreto, C.G. (1972). El lxico de la delincuencia en Puerto Rico. University de Puerto Rico.
Andreini, L. (1985). Le verlan. Petit dictionnaire illustr. Henri Veyrier.
Av-Lallemant, F.C.B. (18581862). Das deutsche Gaunertum, vol. 4. Brockhaus.
Berger, P. & T. Luckmann (1966). The social construction of reality: a treatise in the sociology of
knowledge. Doubleday.

. Address: 38 Rue Nationale, 75013 Paris; tel: 45 84 97 32.

Jargon 169

Beier, L. (1995). Antilanguage or jargon? Canting in the English underworld in the sixteenth and
seventeenth centuries. In P. Burke & R. Porter (eds.): 64101.
Brophy, J. & E. Patridge (1930). Songs and slang of the Britsh soldier 19141918. Patridge.
Burke, P. (1995). Introduction. In P. Burke & R. Porter (eds.): 121.
Burke, P. & R. Porter (eds.) (1995). Languages & jargons. Polity Press.
de Chaves, C. (1983 [1592]). Relacin de la crcel de Sevilla. Clsicos el rbol.
Choron-Baix, C. (1990). Le parler des boxeurs. Jalons pour un jargot. Documents de travail du Centre
dArgotologie vol. 10: 107109. University Ren Descartes.
Coyaud, M. (1990). Largot en Chine. Documents du travail du Centre dargotologie. University Ren
Descartes, vol.10.
Culbertson, R.G. & E.P. Fortune (1986). Incarcerated women: Self-concept and argot roles. Journal of
Offender Counseling, Services and Rehabilitation, 10(3): 2549.
Dauzat, A. (1946 [1929]). Les argots. Delagrave.
Deleuze, G. & F. Guattari (1980). Mille Plateaux. Minuit.
Descamps-Hocquet, M. (1989). Bibliographie des argots franais. Centre dArgotologie.
Dirckx, J.H. (1983). The language of medicine: Its evolution, structure, and dynamics. Praeger.
Foucault, M. (1977). Discipline and punish. The birth of the prison. Penguin Books.
Franois, D. (1968). Les argots. In A. Martinet Le langage: 620635. ditions Gallimard.
Franois-Geiger, D. (ed.) (1990). Parlures argotiques. Langue Franaise 90.
Garca Ramos, J. (1990). Lenguas marginales. Anlisis y vocabulario. Direccin General de polica.
Ghnter, L. (1919). Die deutsche Gaunersprache. Quelle und Meyer.
Giraud, P. (1966). Largot. PUF.
Goffman, E. (1961). Asylums. Doubleday & Company.
(1963). Stigma. Notes on the management of spoiled identity. Prentice Hall.
Halliday, M.A.K. (1978). Language as social semiotic. The social interpretation of language and meaning.
Edward Arnold.
Herman, E.S. (1992). Beyond hypocrisy. Decoding the news in an age of propaganda, including a
doublespeak dictionary for the 1990s. Pantheon Books.
de Klerk, V. (1990). Slang: a male domain? Sex Roles 22: 910, 593.
Kluge, F. (1895). Deutsche Studentensprache. Trbner.
Lafage, S. (1990). Largot the jeunes ivoiriens, marque dappropiation du franais? Langue Franaise
90: 95105.
Lombroso, C. (1876). Lhuomo delinquente. Hoepli.
Louis, P. & L. Primaz (1990). Skinheads, taggers, zulu & Co. La Table Ronde.
Marples, M. (1940). Public school slang. Constable.
(1950). University slang. Williams and Norgate.
Martin, J. (1985/89). Factual Writing: exploring and challenging social reality. Oxford University Press.
Martn Rojo, L. (1989). Anlisis sociolingstico de la jerga delincuente. Ph.D. University de Valladolid.
(1991). Sobre la variacin lingstica. Foro Hispnico 2: 95108.
(1993). De la excepcin al paradigma: anlisis de los fenmenos lingsticos de la jerga
delincuente. In M. Torrione (ed.): 155195.
(1994). The jargon of delinquentes and the study of conversational dynamics. Journal of
Pragmatics 21: 243289.
Martn Rojo, L. et al. (eds.) (1994). Hablar y dejar hablar: Sobre racismo y xenofobia. University
Autnoma de Madrid.
Mausser, O. (1917). Deutsche soldatensprache. Trbner.
Merle, P. (1990). Les blues de largot. d. du Seuil.

170 Luisa Martn Rojo

Michael, S. (1993). Criminal slang in Oliver Twist: Dickens survival code. Style 27(1): 49.
Moncada, S. (1619). Discurso de expulsin de los gitanos. In S. Moncada Restauracin poltica de
Espaa y deseos pblicos. Luis Snchez.
Niceforo, A. (1972 [1897]). Il gergo nei normali, nei degenerati, nei criminali. Forni.
Niceforo, A. (1912). Le gnie de largot. Mercure de France.
OBarr, W. (1982). Linguistic evidence. Language, power and strategy in the Courtroom. Academic Press.
Objartel, G. (1985). Studentische Kommunikationsstile im spteren 18.Jht. In D. Kimpel (ed.):
Mehrsprachigkeit in der deutschen Aufklrung: 2841. Hamburg.
Palmireno, L. (1573). El estudioso cortesano. Valencia: ex typographia Petri a Huete.
Parry, J.H. (19481949). Sailors English. Cambridge Journal 2: 660670.
Patridge, E. (1933). Slang. Today and Yesterday. Routledge & Kegan Paul.
Paz Prez, C. (1994). Diccionario cubano de trminos populares y vulgares. Instituto Cubano del Libro.
Pujolar, J. (1995). The identities of La Pea. The voices and struggles of young working-class people in
Barcelona. Ph.D. University of Lancaster.
Roeding, J.H. (17941798). Allgemeines Wrterbuch der Marine, 4 vols. Nemnich.
Robert LArgenton, F. (1990). Larlepen largomuche du louchbem. Parler largot du boucher. Langue
Franaise 90: 113125.
Sainan, L. (1907). Largot ancien. Champion.
Salillas, R. (1896). El delincuente espaol. El lenguaje. Librera de Victoriano Surez.
Sanmartn, J. (1996). El argot de la delincuencia (Investigacin en el establecimiento penitenciario de
Valencia. Ph.D. University de Valncia.
Sastre, A. (1980). Lumpen, marginacin y jerigona. Legasa.
Singy, P. (1986). Le Vocabulaire medical: jargon ou argot? Linguistique 22(2): 6374.
Sourdot, M. (1990). Argot, Jargon, Jargot. Langue Franaise 90: 1327.
Toro Marzal, A. (1975). Sistemas de investigacin del lernguaje del delincuente. Revista de estudios
penitenciarios 208211.
Torrione, M. (1994). El gitano-espaol: de la etiqueta germanesca a la catalogacin lingstica (siglos
XVXVIII). In L. Martn Rojo et al. (eds.): 95112.
Torrione, M. (ed.) (1992). Lengua: libertad vigilada. CRIC-Ophrys.
Vidocq, E. (1837). Les voleurs. Vidocq.
Wagner, M.L. (1924). Notes linguistiques sur largot barcelonais. Institut dEstudis Catalans.
Yoffe, M. (1992). Russian Hippie Slang, Rock-n-Roll Poetry and Stylistics: The Creativity of Soviet
Youth Counterculture. Dissertation Abstracts International, A: The Humanities and Social
Sciences 52(10): 3621.

Language change
Raymond Hickey

University of Duisburg and Essen

1. Introduction
This overview considers research on language change and the various issues which
have arisen and been discussed over the past few decades. The article considers the
methodologies applied to the field from the comparative method and internal reconstruction to the text corpus and sociolinguistic/variationist approaches of recent years.
Further sections of the article deal with issues such as pathways of change, e.g. grammaticalization, and with explanatory models of language change, such as speakerinduced language change, contact accounts and the typological perspective. The debate
between functionalism and formalism is also given consideration.
It is an obvious truism to say that, given the dynamic nature of language, change
is ever present. However, language change as a concept and as a subject of linguistic
investigation is often regarded as something separate from the study of language in
general. Recent research into the topic, however, has strived to highlight the continual
nature of change and to emphasize that the synchronic and diachronic views of change
can be unified, providing a panchronic perspective in which the relevance of small
changes observed in the present can be shown to hold for larger changes viewed over
longer periods in the past. Furthermore, research in the last three or four decades has
been concerned with understanding the precise mechanisms of change just as much
as with providing linguistically acceptable accounts of attested changes.
The field is served well by literature. There are journals which are dedicated specifically to language change, such as Diachronica, or to certain aspects of this, such
as the Journal of Historical Pragmatics or to questions of methodology which have
a bearing on recognising and documenting change, such as Corpus Linguistics and
Linguistic Theory. In addition, other journals strongly favour analyses that illuminate
language change, frequently from a certain perspective, such as Language Variation
and Change (see also the collection by Chambers, Trudgill and Schilling-Estes (eds)
2002). There are many textbooks that deal with language change (Aitchison 2001;
McMahon 1994; Bauer 1994), frequently under the heading of historical linguistics.
From the English-speaking world one could quote as representative examples Bynon
(1977); Milroy (1992); Trask (1996); Campbell (1998); Crowley (1998); Fennell
(2001); Traugott& Dasher (2002); Beard (2004); Traugott & Brinton (2005); Jones
and Singh (2005); Trask and McColl Millar (2010). Some of these deal with the three

172

Raymond Hickey

main views of change which had been proposed by the middle of the 20th century,
namely the Neogrammarian model from historical linguistics in the 19th century (see
the discussion in Labov 1981), the structuralist approach, initiated by de Saussure at
the beginning of the 20th century and the generative approach, which while beginning in the 1950s as a purely synchronic approach to language description with the
work of Noam Chomsky, came in the following decade to be applied to issues in language change, see King (1969) and other influential works such as Lightfoot (1979).
As generative grammar has undergone many revisions, the analyses of language change
stemming from the standard theory of the mid-1960s have been left aside and more
subtle and nuanced approaches have been developed. In particular, deficient analyses,
which derive from the model used at the time, have been abandoned. For instance,
the scholarly concern of the late 1970s with rule-ordering, particularly in phonology
(see Kenstowicz& Kisseberth 1979 and Koutsoudas 1976), is now seen to result not so
much from the data examined but from a non-linear, unstratified view of phonology,
something which was remedied with the advent of the various forms of syllable-based
and metrical phonology.
The study of language change, rather than the analysis of specific instances of
change, had already been undertaken in the 19th century. Hermann Pauls Prinzipien
der Sprachgeschichte (1880), Principles of language history, shows a linguist standing
back from the monumental task of comparing and reconstructing the many IndoEuropean languages and discussing the underlying principles of this enterprise (Baldi
1991). It came to be perceived as the definitive statement on the historical approach to
language analysis which saw the notion of sound law, German Lautgesetz, as central
and stressed the exceptionless nature of this, German Ausnahmslosigkeit.
It was not until Edward Sapirs Language of 1921 that a major twentieth-century
work reflected specifically on aspects of language change, although Saussures seminal
work on structuralism provided the theoretical framework for all studies until well
into the second half of the 20th century and for many linguists still does. Sapirs most
significant contribution to language change is the notion of drift, an imperceptible and
slow movement in a particular direction which a language can show over centuries
and which can change its typology. Although the notion is controversial and prone to
vagueness, a sympathetic interpretation would see it as an abstraction of the tendency
in each generation to favour certain types of variants present in a language, and importantly, for some speakers to (unconsciously) select the more innovative of these variants.
The latter can then offer a principled account of how drift comes to be observed over
long periods of time. It should be stressed that notions of drift, which lie outside the
data of a language, are vacuous and misleading.
With hindsight one can see that later American structuralism was restricted in its
range, though what it achieved was obviously significant and important for the practical

Language change

work of language recording and description. Leonard Bloomfield, as the major figure
of the inter-war years, neglected semantics, which he saw as a domain of psychology,
and his accounts of language change were mechanistic. It was not until the 1960s with
the work of William Labov, when sociolinguistics was established as an independent
subdiscipline within linguistics, that a theoretical framework for language change with
an innovative approach was presented, see Weinreich, Labov& Herzog (1968). This
has led to a paradigm in linguistics in which scholars extrapolate from small presentday changes in language use to larger attested cases of language change in history.

2. Issues in language change

There are recurrent themes in language change. These have to do with the sources of
change, the nature of the factors involved and, importantly, the significance attributed
to the latter. In the following sections a number of these key issues are presented in
summary form. The factors tend to come in pairs because they represent opposing or
complementary forces. Furthermore, the relative weight accorded to certain factors
varies among scholars. For instance, the question of just what constitutes simplicity
or symmetry in language change is a matter of much debate (see the discussions
in Lass 1997). The role which language contact plays and has played in language
change is also a matter much discussed in recent literature (see the contributions in
Hickey (ed.) 2010).
2.1 Internal and external factors
Perhaps the first division to be made among factors in language change is that between
those which operate from within the language (internal factors) and those which
are active from outside (external factors). These factors are different in themselves
(Gerritsen& Stein 1992). For example, internal factors have very often to do with the
establishment of morphological regularity (analogical levelling and possible analogical
extension, Campbell 1998:9299) or with the reshuffling of items in a word field
(re-alignment of sense relations). External factors have primarily to do with the symbolic
role of language in society. The levels of language first affected are usually phonetics
and phonology, though others may be later subject to change. Internal change is furthermore associated with first language acquistion especially when it involves paradigmatic regularisation and levelling, processes typical of children constructing their
version of the language they are exposed to. In high-contact scenarios (Dorian 1993)
this can apply to adults as well, particularly if they are engaged in language shift in an
unguided, non-prescriptive situation.

173

174 Raymond Hickey

2.2 Simplicity and symmetry

Internal change, on a morphological level, has first and foremost to do with the establishment of regularity within paradigms. This is probably due to the intuitive grasp
which speakers have of word classes and sets of forms within their language. Symmetry
in the linguistic sense would thus be interpreted as a force which seeks to establish
regularity and predictability in the paradigms of a language. If speakers remove irregularity in paradigms across a broad section of a speech community then the resulting
regularisation can lead to language change by being transmitted as default to following
generations. There are many well-known cases of this: for instance, umlaut in Middle
High German spread from nouns such as Haus: Huser to others like Baum: Bume
as speakers assumed similarity in paradigms which showed formal similarity. A similar
pathway of change can be assumed for the spread of the -s plural in the history of
English, replacing earlier -r and -n plurals.
Simplicity, in the sense of a lack of formal complexity, is something which is much
more difficult to quantify in language change. Certainly morphological complexity
does not appear to be a sufficient condition for language change, as complex languages
such as German, Finnish or Turkish amply testify. Perhaps it is better in the context
of language change not to speak of simplicity but rather of the dominance of certain
patterns. For instance, it is questionable whether strong verbs in Germanic are more
complex than weak ones. But the latter are certainly the dominant pattern and in the
history of the Germanic languages there has been a general tendency to replace strong
by weak verbs. Talking of dominant patterns does, however, shift the question to
another level: the reason why a weak pattern has become dominant has not to do with
putative inherent simplicity but rather with the fact that it maintains the constancy
of the lexical stem on inflection. Stem constancy would seem to be valued by child
language learners, because early in the acquisition process they produce weak verbs.
It might be more prudent then not to refer to simplicity but rather to a valued principle
in language organization. The issue of frequency is also important here as this has a
snowball effect: once a pattern begins to be dominant it accrues further instances and
these in turn cause more to be drawn into the orbit of the pattern.
2.3 Iconicity and indexicality
While it is true that the principle of onomatopoeia is not regarded as being of great
significance to modern languages, iconicity is regarded as something which is still visible
(Haiman 1985; Fischer & Nnny 2001). It is important here to distinguish between
two types of iconicity. The first is what is called direct or imagic iconicity and can be
observed where there is an extra-linguistic motivation in the use of certain structural
features which establish a parallelism between linguistic form and what is signified
(Croft 2000:141). A good example of this is the word for soft in Irish, a monosyllable

Language change

which only contains voiced consonants, bog. Equally the word for hard, crua, begins
with a voiceless velar. However, one should not be too adamant about the validity of
this principle, for instance the English word big contradicts iconicity as it has a short
front high vowel, usually associated with smallness, cf. bit.
The second type of iconicity is what is termed diagrammatic iconicity (Haiman
1983) where there is a plurality of signs and where the relationship between these
signs mirrors a similar relation between objects or actions. An example of this can be
seen in syntax when there is a correlation between the linear order of elements and
the sequence of events. For instance, in John hit Bill, John initiated an action which
resulted in Bill being hit. Again while recognising this correlation there is no way it
can be seen as a restraining force on language change. Indeed it is the dominance of
other aspects of syntactic organization, such as topicalization, which has often led to
the rise of alternative word orders, such as VSO. In semiotics an index is understood
as something which points to something else it is closely associated with, e.g. smoke is
an index of fire. In this sense, a certain element can become an index of another and be
preferred over other elements. For instance, word-final /-s/ is an index of a grammatical
inflection in English and probably attained this role from the high phonetic salience of
/s/, hence the preference for it in present-tense verbal inflections and of course in the
plural. Another element which is frequently an index of a grammatical category is final
/-n/ in German which can be the realisation of many verbal endings, weak adjectives,
nasal plurals, etc. It should be stressed that an index in language is a linguistic convention without a causal connection between a sign and what is signified.
2.4 Markedness and naturalness
Among the terms especially appealed to in explanations of language change are
markedness and naturalness. The difficulty with these terms lies in determining exactly
what they refer to. The matter is compounded by the fact that the terms are frequently
not defined. For the present discussion the term marked is taken to apply to an element
which is cross-linguistically unusual, i.e. statistically rare in the worlds languages. This
fact of course leads to the question why this is the case. For example, dental fricatives
do not occur very frequently in the languages of the world. The reason for this may
have to do with their low acoustic salience, especially when compared to sibilants.
However, the reason for the rise in frequency of such unusual segments may well be
linked to a very common process itself. Here intervocalic lenition can lead to dental
stops becoming fricatives as happened historically in Spanish. This highlights another
aspect of the term marked: it is inherently comparative, that is it is a statement about
the relative significance of a feature in one language when compared to its possible
occurrence in a larger set of languages. But there is also local markedness, a feature can
be unusual within the context of a single language. The nasal ending as in ox: oxen is

175

176 Raymond Hickey

highly marked in English, however, in German this nasal ending is much more frequent,
cf. die Frau: die Frauen the woman, the women.
The term natural is even more fraught with difficulties. A process can be said to be
natural when it is not unexpected. Judging what is unexpected of course rests on the
experience of the individual linguist and this factor in language change is even more
difficult to quantify than markedness. One objective yardstick for measuring naturalness
could be based on correlations and group affiliation. For instance, it is common for a
language which has the word order VSO to have adjectives following nouns and the
genitive following the nominative. These correlations have to do with the principle of
post-specification. The recognition of group affiliation has led to the notion of natural
class. This can be seen quite clearly with phonological segments: the set of voiceless
stops forms a natural class as all its elements share a lack of voice and an interruption
of the airflow and are typically found in syllable onsets.
The correlations one finds between sets of elements has been used as a basis for
postulating implicational universals, statements about the structure of languages
which seem always to hold. For instance, if a language has voiced stops (B), then it will
always have voiceless ones (A), or if a language has nasal vowels (B) then it will have
oral vowels (A). The justification for such implicational universals is seen in the fact
that in the development of languages over time, elements of the (B) sets are derived
from those of the (A) sets.
2.5 Telic changes and epiphenomena
A major issue in language change is whether instances of it can be regarded as therapeutic for a language, that is whether the change can in any way be seen to serve
a certain purpose within the system. In any consideration of this kind there is an
inherent danger that the language, as a convenient abstraction over the behaviour of
speakers, is seen as taking on a life of its own and motivation for change is imputed
to it. Again if one adopts the speaker perspective, then many instances of change which
appear to be telic (goal-oriented) in fact turn out to be cases of reanalysis by language
learners. To illustrate this, consider the functionalization of initial mutation in early
Celtic. This is a set of far-reaching phonological changes whereby grammatical distinctions, formerly realised by suffixal inflections, came to be indicated by changes in the
initial segments of words, e.g. the change from stop to fricative to indicate third person
possession in the masculine: cuid /k-/ amount, part, a chuid /x-/ his part. On closer
inspection this development turns out not to be a telic change but the result of reanalysis
by child language learners. What appears to have happened is that they interpreted
phonetic weakening at the beginnings of words as systemic, i.e. they regarded these
initial changes as exponents of grammatical categories (Hickey 2003a). The effect of
this was to retain a means for indicating grammatical categories. However, it would

Language change

be erroneous to imagine that speakers initiated this shift in order to reach the goal of
establishing initial mutation.
If one accepts that language change is largely atelic (not goal-oriented), then
the question arises as to whether it is indeed epiphenomenal, whether the observed
instances of language change were not intended but resulted from related alterations
in the language. The genesis of new forms of language from a dialect mixture situation would appear to confirm this. The processes from the speaker perspective may
be dialect levelling and accommodation, but the net result may be the rise of a new
variety of a language at a new location, as happened with many overseas forms of
English during the colonial period. There is much debate on these issues (Hickey
2003b) and an important consideration is whether speakers are aware of their own
linguistic behaviour, at least unconsciously. If this is true, at least in some cases, then
instances of language change which seem to be epiphenomenal could be due to speakers
moving along an unseen trajectory of change.
2.6 Mergers and distinctions
Much along the same lines as the previous considerations is the question whether
language change can be systemically dysfunctional. The question here is really how
to determine if something is dysfunctional. There are many attested cases of mergers
which might appear to be detrimental to communication. However, this is only so at
first sight. The context of any utterance is nearly always sufficient to disambiguate so
that considerations of homophony can only in the rarest of cases be postulated as a
reason for change occurring or not occurring.
Another separate but related issue is whether a merger, once it has occurred, can
be reversed at some later stage. There would appear to be cases of this in the history
of English. For instance in London in the 16th century the vowels in the MEAT and
MATE lexical sets were both /:/ but later these turn up as MEAT /i:/ and MATE /e:/
respectively, more or less their values in present-day English. But as scholars such as
William Labov and James Milroy have pointed out (Labov 1994:349390; Milroy&
Harris 1980), one may be dealing with near-mergers, or if indeed with mergers, then
not perhaps for the entire population speaking a language. There may be varieties
present which did not undergo a merger and which offered a model for the later
re-establishment of the distinction MEAT and MATE after the 16th century.
Even where a systemically important merger has taken place, or significant elements of a language are lost, there does not of necessity have to be a typological
realignment. For example, in Old Irish the loss of phonological length meant that
the contrast of geminate and simplex sonorants was no longer available. Nonetheless, the language did not compensate for this loss of grammatical exponence, that
is speakers did not consider the resulting ambiguity to be serious enough to force

177

178 Raymond Hickey

them to evolve new strategies for the indication of the grammatical categories now no
longer formally evident.
2.7 Possible changes
Closely connected with the question of merger and possible remedial action is the
question of what constitutes an impossible change in language, indeed whether this
can be specified or not. Inherent constraints would seem to be connected to our cognition, e.g. there appears to be no language which allows more than triple embedding
(if at all) of the type The mouse the cat the man owns chased ran because such sentences
are very difficult to process mentally. Other types of embedding, such as that of relative clauses before the nouns they qualify, are possible, excesses being constrained by
short term memory, cf. German Diese lang berlegte Frage, lit. this long considered
question, cf. Diese Frage, die lange berlegte wurde This question which was considered
for a long time.
Cases of change where the starting and the end points lie far apart may at first
sight be unexpected but generally the intermediary steps form a recognisable and fairly
usual sequence of steps. Thus Latin centum and Modern French cent /s/ hundred or
Indo-European gwena and Modern Irish bean /ban/ woman may look far apart but
can be clearly shown to be directly related. There are only a few changes which do not
have intermediary steps, metathesis being one of them, e.g. English kitchen /-tw-/
Irish cistn /-wt-/.
By and large one can say that unusual changes are possible if they are carried by
the speech community, i.e. if they are markers of the community, have high indexical
value and are hence maintained by speakers. Tightly-knit communities with strong
internal networks are good bearers of statistically rare changes (Andersen 1988).
2.8 Unidirectionality of change
Apart from types of changes, attested and unattested, another consideration is whether
the direction which change takes is only one-way. If unidirectionality existed, it would
impose a strong formal constraint on language change. Here, as with universals in
language, one must distinguish between absolute and relative unidirectionality. Among
the contested types of unidirectionality, which have been recently suggested, is the
notion, put forward by William Labov, that peripheral vowels, e.g. /i:, e:, o:, u:/, become
more open and non-peripheral vowels, e.g. /i, , f, ~/, become less open (Labovs Vowel
Shift Principle, 1994:13ff.). This is a principle with caveats such as the Mid Exit Principle
which states that peripheral vowels, rising from mid to high position, develop inglides,
thus accounting for the development of the English long vowel system since the Middle
English period but with diphthongization of mid (and to some extent high) vowels in
the late modern period. The contention that non-peripheral, i.e. centralized, vowels

Language change

become less open is borne out by the current Northern Cities Shift in the United States
and the raising of short vowels across the Anglophone Southern Hemisphere but not
supported by the lowering of short vowels in Scots: /i/ /e/ as in stick [stk].
Unidirectionality is also an issue in grammaticalization theory (see discussions in
Fischer 2000; Kim 2001; the contributions in Wischer & Diewald (eds) 2002). From
the data presented in various studies is would seem that the path which leads to lexical
elements becoming more grammatical in time is largely unidirectional (Norde 2009:
Chapter2). There are few clear cut exceptions, perhaps the Irish example of muid we
which derives from suffix for the first person plural, -m(u)id, as in baileoimid we will
gather. Such cases do not seem to diminish the overall unidirectionality of grammaticalization (Lass 2001; van der Auwera 2002). But there are related phenomena which
do move in the opposite direction, see the discussion of deinflectionalisation and
debonding in Norde 2009: Chapters 5 and (6).
It should be mentioned too that the ontological status of grammaticalization has
been contested in recent years (for critical discussions see Janda 2001 and Joseph
2001). Indeed some linguists have come to believe that it is an epiphenomenon, resulting
from other changes such as cliticization, loss of semanticity, increase in grammatical
categoriality (see the discussion in Fischer & Rosenbach 2000).
2.9 Ebb and flow
When observing instances of language change, not only does one see that many are
not unidirectional, but there would seem to be a movement backwards and forwards
across several generations. A label given to this type of change is ebb and flow (see
Hickey 2002c for a detailed discussion). It can be clearly illustrated by an example
from the history of English, namely that of velarized /R/. There is no doubt that a
velarized (dark) [R] has existed in the history of English and has played a significant
role in determining the phonological form of words containing it. Already in Old
English breaking (diphthongization) occurred before clusters containing /l/ followed
by an obstruent as in eald old. This allophone continued in Middle English and was
later vocalized in syllable-final, pre-obstruent position, especially before velars as
words like talk and chalk in their modern English pronunciations indicate. Much dialect
evidence from England, Scotland and Ireland points to a velarization before alveolar
obstruents as well.
In the present-day English of London and the Home Counties a syllable-final
velarized [R] is very much in evidence and so the conclusion that this represents historical continuity might seem straightforward. But there is evidence that velarized [R]
in syllable-final position was not transmitted in an unbroken fashion for all south,
south-eastern dialects of British English. Trudgill (1999) quotes support from the
Survey of English Dialects (Orton & Halliday 1963) for the view that in large parts of

179

180 Raymond Hickey

southern Britain a velarized [R] is a recent phenomenon going back no further than the
late 19th century. Thus despite the considerable historical evidence for velarized[R] in
Old English (West Saxon) and in the east in Middle English there must have been a
swing of the pendulum away from velarization in the late modern period with a fairly
recent reinstatement of this secondary articulation which is so prominent in varieties
of southern British English today.

3. Change and levels of language

Language change offers evidence for the autonomy, but interconnection, of linguistic
levels. While there is general consensus on the organization of language as an arrangement of subsystems, there is equal insistence on the fact that these subsystems have
definite points of connection. Although the sound and meaning systems of a language
are separate there are definite links between the two, for instance, if there is phonological
attrition in a language then this may well lead to the opacity of lexical forms which in
turn can lead to later reinterpretation as with the many instances of folk etymology
which exist, e.g. sand-blind from samblind half-blind (OE sam- = half ) or bridegroom
from brydguma bride-man (OE guma = man).
Before looking at the individual levels of language it is important to point out
that in recent research linguists have stressed the gradience of grammatical categories
(Aarts et al. (eds) 2004). In English, for example, there is a common practice of using
nouns in attributive adjective position which leads functionally to a change in class,
e.g. with fun (noun) as in A fun party or key (noun) as in Some factors are key. This
is still within a single level of language, that of syntax, but it does show that there is a
degree of fluidity among categories, something which is long established in English
and is perhaps characteristic of analytical languages as a whole.
3.1 Phonological change
On a purely phonetic level one could say that sound change has to do with an increase
or decrease in sonority. Segments are usually strengthened in syllable-initial position
and weakened in medial and final positions. It is one of the great merits of non-linear
phonology to have offered a formal framework for describing these types of change,
that is developments which are dependent on syllable structure and syllable position.
The natural tendency to increase the sonority in medial and final syllable positions
has led in many languages, for example in the Indo-European daughter languages, to
a loss of inherited inflections something which has over several centuries led in many
cases to a typological realignment of individual languages, e.g. with the Romance languages vis--vis Latin or English vis--vis the Germanic parent language with analytical

Language change

structures becoming generally dominate in such instances. Grammatical reorganization

due to phonological change illustrates quite clearly the interface between the sound
and grammatical levels of language.
Another common phonetic development is the exploitation of phonetic polarizations. For instance, if glides become absorbed into preceding consonants, then the
effect can be palatalization (with later reactive velarization) as in Q-Celtic and Slavic,
to mention only two of the more well-known cases. A similar functionalization of
polarization can be seen in the development of emphatic consonants in Arabic.
Phonetic changes of the type just described are gradual: but the ultimate effect
can be a reanalysis of the exponence of grammatical categories in a language. Two
exceptions to gradual phonetic change are epenthesis and metathesis, the introduction of vowels/consonants in specific clusters and the change in the linear order of
segments in a word respectively. However, neither of these processes appears to have
attained a grammatical function in a language. The reason may well be that a language
either has epenthesis and/or metathesis or it does not. But for grammaticalization to
occur there must be the option of contrast in a language and this would seem to be
precluded with the two processes just mentioned.
3.2 Morphological change
For many languages morphology is the mediator between deep syntactic case and
surface realization. It is obvious that there is no language without syntax and phonology
but there are languages with little morphology (prototypically analytical languages).
This fact might seem to suggest that morphology is not part of universal grammar,
at least that it can remain unrealized. Surface morphology can be seen as the result
of performance phenomena such as phonetic attrition, leading in the fullness of time
to cliticization and grammaticalization. It is identified as a level of language as it is
manifest on the surface and indeed there is a sense in which, in those languages with
morphology, that it is a level which can be given a ranking above phonetics.
Although the ultimate status of morphology can be debated (because it is not
necessary for human language), it is nonetheless a reality for all language types apart
from analytical languages. In the historical development of languages it is probably
true to say that morphology has always arisen through the contraction of independent elements. For instance, suffixal inflections can be seen to arise where, in a word
dyad, one element is a lexical stem and the other a form word. The latter can go
through a process of semantic bleaching and cliticization (attachment to the stem)
and ultimately become an inflection, i.e. lose its independent phonetic form and no
longer be detachable from the stem. An example would be the unstressed ending -dom
[dm] in kingdom, wisdom, freedom which is related to OE dm in the meaning of
condition, domain.

181

182 Raymond Hickey

Once a language has developed a morphology certain forces become dominant

which involve specific pathways of change. Nowhere is the operation of internal factors
more in evidence than with morphology. This probably has to do with the fact that
morphology leads to the organisation of words into sets of forms and that for speakers,
particularly for language learners, regularity within such sets is highly valued. Irregularity may arise for a variety of reasons, for instance due to suppletion, the mixture of
elements from two or more sets within a single paradigm. Internal irregularity within
a single paradigm may arise due to certain phonological factors. A good example of
this is the voicing of segments where the immediately preceding syllable does not carry
stress, e.g. executive [igzekju:tiv] versus execute [ekskju:t] in present-day English,
which in the Indo-European context is known as Verners Law (Trask 1996: 226f.).
This together with the shift of /s/ to /r/ (rhotacism) led to considerable irregularity
within nominal and verbal paradigms in Germanic. At a later stage analogy (Trask
1996:105115), the spread of a dominant pattern to an entire paradigm the internal
factor in language change par excellence came to bear (Lahiri (ed.) 2000). Hence in
English one finds a sibilant in the verb lose whereas in German /r/ is found throughout
the corresponding verb, verlieren. However, in Dutch one has the infinitive verliezen
(with a sibilant) to lose but the simple past has an /r/: verloor lost.
Morphology is rarely involved in externally motivated change. The reason probably
is that the latter is driven by social factors, and is dependent on speakers making
choices, on inherent variability. This is the case with the phonology of a language: to
put it in simple terms one can pronounce sounds/words in different ways and so give
social significance to this variation. Equally in syntax one usually has a number of constructions which are roughly equal in semantic terms and so again one has choice. But
the system of morphology is much more rigid. However, in those cases where there
is variation, for instance in verbal tense marking or conjugation in English (Ihalainen
1994), then this variation can become relevant for language change, frequently leading
to the differentiation of varieties of a language.
3.3 Syntactic change
In an overview chapter such as the present one it is impossible to do justice to the
vast theme of syntactic change (Roberts 1993; Harris & Campbell 1995; Kemenade&
Vincent 1997; Pintzuk, Tsoulas & Warner 2001; Fischer 2006; Roberts 2007; Crisma&
Longobardi 2009; for greater time depth see Heine & Kuteva 2007). All that one can
do in the current context is to point out some general features of the phenomenon and
advise readers to consult the relevant literature for detailed discussion of instances and
interpretations of syntactic change. It should also be mentioned that syntactic change
tends to be the domain of linguists with a more formal orientation. Functional views
of language change are often associated with a sociolinguistic or typological approach

Language change

to the field (see the discussion in Newmeyer 1998, 2002). Within the possible formal
approaches recent scholars have often contrasted grammaticalisation and generative
standpoints with each other, see Fischer (2006: Chapter 3).
There is an essential difference between phonological and syntactic change in
that syntactic variables do not occur as frequently in speech as phonological ones.
Syntactic variation is more likely to be conditioned by internal linguistic factors or
depend on questions of style and context rather than external social factors since
syntactic structures are repeated less often than phonological ones and are thus less
available for social assessment. Furthermore, syntactic factors are less visible as the
structures which they engender are more abstract and hence speakers are less conscious of them.
Syntactic structures do not usually have a social identification function like phonological factors as there may well be stretches of speech in which a given syntactic variable
does not occur at all and hence does not offer a speaker a clue as to the linguistic affiliation of an interlocutor.
A corollary of this fact is that emphatic/prominent contexts, such as pre-verbal
positions, main clauses, topicalized declaratives and explicit negatives, would appear to
be favoured sites for variables which have a social significance in a community. Conversely, less prominent contexts may be the sites at which incoming variants may first
appear because of their low salience and spread from here to more marked contexts. So
there may well be two types of context: (i) a high profile one used for variables which
have high social significance by speakers who identify strongly with a community and
(ii)a low profile one used for variables which are incoming and being adopted by diffusion into a community. This second context is important for views of language change:
change may at first be in evidence in less common contexts, i.e. it might appear as if less
frequent patterns are at the leading edge, contradicting a notion that the more common
pattern is the one which leads the way. Typical sites with such a low profile would be
post-verbal positions, subordinate clauses, weak affirmatives (Cheshire 1996).
If low profile sites are preferred points of entry for incoming variants this implies
that there is a donor community which already has the variants in question in high
profile sites. But for various reasons, such as attestations in history, evidence of the
high profile primary community might be lacking and the status of low profile sites
as entry points for incoming variants may not be evident to later investigators of a
language/variety.

4. The study of universal grammar

The remarks so far have all referred to externally triggered syntactic change. However, for many linguists the primary concern is determining the internal constraints

183

184 Raymond Hickey

on change. The ultimate goal of such an enterprise is to provide clear contours to the
notion of what constitutes a possible human language and, in the historical context,
what transitions occur between language states. This quest is generally known as the
study of universal grammar and in the last three decades or so much progress has
been made in the formulation of constraints on the structure of human language.
These constraints are frequently couched in terms which are only understandable
within the framework of generative grammar as it developed through the 1980s as
can be seen from the following four examples (though example four is more accessible
to the non-initiated).

(1) Subjacency: A moved element cannot be separated from its trace by more than one
binding element.
(2) Tensed S constraint: No rule involves two elements respectively inside and outside
a tensed S.
(3) Disjoint coreference: No two NPs may be interpreted as intersecting in reference.
(4) Cross-categorial generalization: Verbs and prepositions take formally similar
arguments (traditionally accusative case).

For reasons of space details of universal grammar cannot be discussed here and hence
it is not possible to show the many insights into language structure which its study has
resulted in. Instead one aspect of this complex, namely change in language centred
around the language acquisition process will be dealt with briefly to illustrate the type
of thinking which is characteristic of formal linguists with regard to language change.
4.1 The principles and parameters model
Among linguists of different theoretical persuasions there is much debate about the
locus of language change. For sociolinguists this lies in the inherent variation in language used by adults, especially on the phonological level. For formal grammarians the
locus of change is in childhood during the acquisition process (Fischer etal. 2000:5).
The reason for this is that language learners construct their internalized grammar from
the performance of those surrounding them. In the model known as principles and
parameters (Lightfoot 1991; Fischer 2006: Chapter 3), linguists assume that children
begin language acquisition from an initial state of universal grammar which contains
abstract, non-specific structural information about human language and then move
during the acquisition process to a state characterized by grammar, where various settings for parameters of language have been reached, going on the values in the language
which the children are exposed to. Linguists working within this framework may have
differing views according to the role they ascribe to universal grammar. Clark & Roberts
(1993) assume that language learners match input data from their environment with
all possible parameter settings permissible by universal grammar. They proceed to

Language change

construct fragments of their grammar and combine existing fragments to larger ones,
weighing up competing subgrammars until a final one is reached which then becomes
their adult grammar.
As children obviously have no direct access to the internalized language of the
preceding generation, typically as exemplified by their parents, their grammar must
of necessity be an approximation to, and not a replica of, the internalized language of
their predecessors. For some linguists, for instance for Lightfoot (see Lightfoot 2002)
the construction of grammars is highly sensitive to the initial conditions under
which children initiate the process, i.e. it is chaotic in the technical sense, and so
there is no reason to expect to find a predictive theory of change, offering long-term,
linear predictions.
A further point stressed by other formal linguists is that changes frequently
come in sets of related instances, as Roberts (1985) has shown in his analysis of the
development of English modals which was offered as an alternative to that proposed by
Lightfoot (1979) which was based on his transparency principle by which children seek
to create the simplest path in the derivation of surface structures from underlying ones.
Related to the transparency principle, but not identical, is the earlier notion of
abduction (Andersen 1973) which states that children reach conclusions about the
structure of the language they are acquiring which are not logical consequences of
premises, but nonetheless made. An instance of abductive reasoning (from morphology)
would be the following: the majority of words ending in -nis in German are neuter, a
child thus assumes that a new word, like Erlaubnis, is neuter, but in reality it is one of
the few instances of feminine gender. Another instance can be seen in North and West
Germanic: the co-occurrence of a change to high front vowel with certain inflectional
endings (umlaut) came to be interpreted by language learners as the exponent of the
grammatical category in question and the inflections became increasingly indistinct
and eventually lost, certainly in English and to a considerable extent in the North
Germanic languages.
There has been much criticism of views of language change which see it as determined during the acquisitional process. This assumption implies a certain abruptness to
change, a parameter is altered once and for all and after that the language has changed.
But surface attestations of languages tell a different story, one of graduality and variation which leads imperceptibly from one language state to another through grey transitions. A theoretical discomfort with the binary notion of language change has led
other linguists, see for instance the work of Pintzuk (e.g. 1991) or Kroch (e.g.1989),
to assume that speakers, for many fragments of their adult grammars, have more than
one parameter setting. This is a type of bilingualism in which two or more competing
subsystems are accommodated within the competence of one speaker. Other models
which strive to avoid such abrupt views of change are those which stress the gradience
of change (Denison 1999; Aarts et al. (eds) 2004) or the gradual spread of change to

185

186 Raymond Hickey

encompass all possible inputs (in the lexicon), a view captured by the label lexical
diffusion (Wang 1969; Chen & Wang 1975; Phillips 2006).

5. Semantic change
Change on the meaning level of language is largely dominated by two forces: metaphor
and metonymy (Kittay 1987; Mac Cormac 1985; Barcelona 2000). The first, metaphor,
has to do with non-literal uses of words. The second, metonymy, is a type of change
where new meanings arise from something closely related to another meaning present in the discourse, frequently within the relationship of part to whole. Metaphorical
change can be seen in polysemy where a figurative meaning of a word develops alongside a more literal one, for example at the foot of the mountain from the primary meaning of foot as part of the body at the base. Polysemy can be said to exist as long as the
primary meaning is still available. If this is lost one is left with an instance of metaphorical change, e.g. the word portfolio flat case contains folio sheet of paper deriving
from the Latin word for leaf through Italian. In fact the metaphorical change has been
continued with the more recent senses of (i) range of investments or (ii) proof of suitability for employment. Metaphorical extensions can equally be found in phrases and
not just single words, e.g. What flows from that? What are the consequences? To wear
the blame for something To accept responsibility, Demand a level playing field Insist
on fair terms.
Metonymic change can be seen where associated or partial elements of a meaning
become the main bearers of meaning, e.g. Cologne as the word for perfume or aftershave from Eau de Cologne. The textbook example of metonymic change is the use of a
city name for a government, e.g. Berlin is not interested in adopting the new proposals.
London has turned down the offer. Metonymy can be seen to operate on other levels
of relatedness, for instance where the meaning of a word changes to a closely related
sphere or process, e.g. English chase (from Central French) indicates a process and
catch (from Norman French) refers to the result. Apart from process and result, other
typical metonymic relationships are (i) action and instrument as in The nine fifteen
service to Edinburgh, i.e. the train which provides the service, (ii) space and time
connectivity, e.g. German Dasein existence which derives from da sein to be there
[locative], to exist.
Metaphorical and metonymic shifts are types of changes. There are others which
are connected with scope or evaluation. Semantic widening and narrowing are commonly attested changes, e.g. English joy is presently being expanded to include success,
students now include pupils. A common motivation for semantic narrowing is social
decorum, hence the restriction in the use of the word intercourse because of one of its

Language change

uses in the sexual sphere. Semantic amelioration is found where the meaning
of a word becomes more positive, again a textbook example is nice which over the
centuries has moved from ignorant (< Latin nescius not knowing) to foolish to
harmless to pleasant. Semantic pejoration can be seen with silly which derives from
Old English blessed (cf. German selig or Dutch zalig).
For reasons of space it is not possible here to discuss the various means and
devices for creating new words. A good overview of these is offered in Campbell
(1998:273279). For general overviews of semantics, see Wierzbicka (1996) and the
contributions in Lappin (1996). In addition it should be mentioned that in recent years a
comprehensive literature has developed within cognitive grammar which is concerned
with semantic change. See Allwood & Grdenfors (1999) on semantics and for more
general discussions, see Dirven & Verspoor (1999); Heine (1997); Jansen & Redeker
(2000); Langacker (1991, 2000); Ungerer & Schmid (1996).

6. Pragmatic change
As the study of language in use, pragmatics covers a wide and not always tightly structured area. In recent years it has been very much the focus of linguistic attention and
specifically its historical dimension has been developed (Jucker 1998). In terms of
language change one can note shifts of semantic meaning to pragmatic meaning, that is
an increased reliance on contextual information in a number of cases. A clear example
of this is provided in Irish where concessive clauses came to be expressed with the
simple clause conjunction agus and (this has also been carried over to Irish English)
Chuaigh s amach agus ag cur baist. lit. He went out and it raining (Irish English),
i.e.He went out although it was raining. Similar instances of reliance on context are
found elsewhere, e.g. German wenn = if and when, with the context preventing
ambiguity: Er macht die Aufgabe wenn er Zeit hat. He will do the task if/when he has
time (there is also falls meaning if which can be resorted to if necessary).
If pragmatics is taken to refer not just to context but also speaker perspective
then it can also be seen operating in another type of change, namely where the subjective view of the speaker more and more informs the use of a word or construction
(Traugott 1989). Meanings tend to become increasingly rooted in the speakers subjective belief, state or attitude toward what is being said as with the discourse marker
use of such adverbs as after all or the performative uses of locutionary verbs such as
promise, recognize. The standard example of subjectification is (concessive) while in
English (While she is interested in linguistics, she nonetheless chose literature) deriving
from Old English dative hwilum at times. For a detailed discussion of semantic change
and the principles and forces which underlie it, see Dasher & Traugott (2002).

187

188 Raymond Hickey

The level of language with which pragmatics intersects most often is semantics,
but it does relate to others. For instance, the vocative is traditionally regarded as a case
and indeed has morphological exponence in classical languages such as Latin and
Greek. However, it is largely a pragmatic category which is not as tightly integrated
into the internal structure of a language as are other cases such as the genitive or dative.
Its function is that of calling attention and has been preserved in many modern IndoEuropean languages, like Irish, for just this purpose.
If pragmatics is given a comprehensive interpretation then it can also be seen to
encompass features of discourse organisation in a language. Changes can take place
in a language, or varieties of a language, which results from the demands of discourse.
An example to illustrate this is the maintenance of a formal distinction between singular and plural personal pronouns in many non-standard varieties of English after its
demise in standard English. This can be seen in the use of ye, yez, youse, yall, youuns,
etc. to express plurality with second person pronouns (Hickey 2003c).
The relationship of pragmatics and syntax is an issue which has been the subject of increasing scholarly attention in recent years (see the discussion in Fischer
2006:210257), after an early distinction between a pragmatic and a syntactic mode
was introduced by Givn (1979:223). Items of change from well-studied languages,
such as English, have been (re-)interpreted in the light of the interplay between pragmatic and syntactic factors. An instance would be the development of verb complements
consisting of that-clauses through to-clauses to bare verb stem as in the following (see
Hopper& Traugott 2003:179).
I think that he should be more careful.
We wish to see him back on his feet again.
She saw the old lady fall.

No single analytic model can claim to account exhaustively and accurately for the data
and as Fischer points out (2006: 248) apparent grammaticalisation pathways, with
clause reduction leading to fusion, are not always upheld by the historical record. Other
factors such as frequency and entrenchment as well as adjacency and analogy effects
may have been equally if not more important in such diachronic developments.

7. Methodologies
The discussion thus far has been concerned with attestations of language change. But
what scholars register as language change depends crucially on the methods used. Traditionally there are two main methods, (i) the comparative method (Durie& Ross 1996;
Campbell 1998:108162) and (ii) internal reconstruction (Campbell 1998:201225).

Language change 189

Both of these are legacies of 19th-century linguistics when the methods were developed
and optimized for research into the Indo-European languages.
7.1 Comparative method
The comparative method rests essentially on comparing two or more languages in
the attempt to extrapolate backwards and postulate earlier forms common to these
languages. It is a legitimate heuristic given certain provisos. It rests on the assumption that there are regular correspondences between languages, and the observation of
these led to the formulation of sound laws in the 19th century. These in essence stated the
regular correspondences across stages of languages. Because sound laws are not without exception the value in comparative linguistics is relative. However, by considering
other factors, such as the internal factor of analogy, one can arrive at a fair degree of
certainty with regard to earlier postulated forms of non-attested languages. Thus by
looking at English three, thou, thin, German drei, du, dnn and Latin tres, tu, tenuis
one can find evidence for the Germanic Sound Shift which led to inherited /t/ of
Indo-European becoming // in the early phase of Germanic. Looking at the German
data also shows that German (along with Dutch and the North Germanic languages)
later fortified and voiced the dental fricative, hence du is the form cognate with English
thou and Latin tu in the languages just mentioned.
The comparative method requires that one take various, possibly conflicting
developments at different stages of languages into account. For instance, present-day
Danish has an ambidental fricative, despite the fortition of Germanic //, but this is
the result of a later lenition of stops in positions of high sonority, cf. Danish mad /ma/
food, Swedish mat /mc:t/ food, English meat.
The importance of chronological sequencing in change was recognized early on by
historical linguists. But it is frequently difficult to determine the point in time at which
a change occurred by using the comparative method. Instead of trying to attempt an
absolute chronology linguists are very often content with relative chronology, i.e. with
saying that, of two changes or events, one preceded or followed the other. For instance,
in Old English one can say that umlaut followed palatalization so that the word cyning
king from /kuni/ did not have a front vowel at the time palatalization was active
and hence did not become /tyni/. Another instance is English blood from Middle
English /blu:d/. With the lowering of /u/ to /%/ in the south of England in the 17th
century this word was altered to /bl%d/ because the /u:/ had been shortened beforehand. In the case of took the shortening occurred much later and was unaffected by the
lowering of /u/ to /%/, hence the pronunciation /tuk/. Relative chronology can also be
useful when dealing with borrowings, for instance, English wine is a Latin loanword,
vinum, borrowed in continental Germanic when Latin v was /w/. The word vine is a later

190 Raymond Hickey

borrowing of the same word in the Middle English period from Latin via Old French
where the pronunciation of v was /v/.
7.2 Internal reconstruction
In its type of argumentation, internal reconstruction is similar to the comparative
method, but applied to a single language (Campbell 1998:201). It uses similar techniques, by comparing sound changes, by looking at the operation of analogy and in
general by considering the likelihood of certain processes to have taken place. One of
the best uses of internal reconstruction is to explain present-day irregularity, that is
alternations which are synchronically unmotivated. An example would be stem-final
voicing of fricatives in plurals in present-day English as in knife: knives, wife: wives.
This goes back to an intervocalic voicing rule of Old English which has long since
become inactive but whose reflex can be recognized today in these plurals.
7.3 Analogy
When applying the comparative method and/or internal reconstruction the operation
of analogy is often to be observed. Traditionally, two types of analogy are recognized
(Campbell 1998:90f.). Proportional analogy is operative when a fourth term is created/
altered on the basis of the second term of a four-term formula, e.g. drive: drove::
dive: (dived >) dove. The second type of analogy is non-proportional and has various terms such as analogical creation or analogical levelling depending on what is
involved (Trask 1996:106109). If analogy leads to a new form then one is dealing
with analogical creation, e.g. female is a word created in analogy with male. The original French word, its source, is femelle. Analogical levelling can be seen where diverse
terms in a series are made to conform to a dominant pattern. For instance, in Irish
there is a process called nasalization where after certain numerals a nasal is prefixed to a word, e.g. dn castle but seacht ndn seven forts. This process was triggered
by the numerals 7, 9 and 10, which originally ended in a nasal, cf. Latin septem, novem,
decem. However, the number 8 also causes nasalization in Irish although it did not
originally end in a nasal, cf. Latin oct. However, the analogy with 7, 9 and 10 meant
that it was brought into line with the numbers preceding and following it, hence acht
ndn eight forts.

8. Sociolinguistic investigations
The relevance of sociolinguistics to the enterprise of language change has been repeatedly stressed by scholars in the past few decades. It rests on the assumption that change

Language change

as it takes place today is no different from change as it took place in history. Behind
this belief is the view that the structure of human language and the use of language
in society has not altered to any considerable extent throughout recorded history
(the uniformitarian principle, Lass 1980). Going on this assumption, one can use the
present to explain the past (Labov 1977). This is the basis of panchronic approaches
to language change.
In its methodology sociolinguistics has been instrumental in introducing new
rigour to data collection, specifically in the insistence on random selection of informants (quite different from traditional dialectology) and on the backgrounding of the
linguist to minimize interference by him/her with the presentation of data by informants (the avoidance of the observers paradox to put it in Labovian terms).
Many insights from sociology came in the 1970s and 1980s to be applied to sociolinguistics. Suffice it to mention two of these in the present context: (i) social networks
and their relevance for linguistic behaviour and (ii) accommodation of speakers to
their interlocutors in social settings. Social networks in the work of James and Lesley
Milroy have helped to explain the internal cohesion of varieties and the survival of
forms of speech much removed from supraregional and/or standard varieties which
might co-exist in a society. These two scholars have continually stressed that the locus
of language change lies with the members of a speech community (speaker-induced
language change, see Milroy 2002). Accommodation, above all in the work of Peter
Trudgill, is assumed to account for the rise of mixed dialects through the contact of
speakers from different backgrounds (Trudgill 1986). Individual accommodation is
cumulative and on a macro level it is responsible for the appearance of compromise
dialects such as those which arose due to contact in East Anglia (Britain 1997).
For sociolinguists the result of change is less of a concern than the course which
change can take. Specifically, they have been concerned with the three stages of
(1)actuation, (2) propagation and (3) termination. Sociolinguists have stressed that
certain strata of society are particularly active linguistically and likely to initiate change,
e.g. through hypercorrection in the lower middle classes striving upwards socially. The
spread of a change is characterized by a slow-quick-slow rate, i.e. it begins slowly, goes
through a rapid middle phase and slows down again towards the end (this course can
be represented in visual terms as an S-curve, Denison 2003). The slow end of a change
means of course that not all possible input forms for a change may be reached and a
residue may remain. For instance, in early modern southern English there was a general lowering of /u/ to /%/, cf. cut /k%t/. However, this did not apply to all words, that
is, the change did not go to completion. It fell short, so to speak, and a small residue
remains, as can be seen from words like pull or bush. Frequently there is a phonetic
motivation for such a residue, in this case the unrounding of /u/ (and its later lowering)
was inhibited by a following velarized -l [-R] or -sh [-].

191

192 Raymond Hickey

8.1 Data collection methods

In diachronic investigations the data used has come to be much more robust through
the use of large collections of texts available in electronic form. The practice of compiling
structured texts into coherent groups, often representing different authors, genres and
periods, goes back to the 1960s. But it was not until the early 1990s with the widespread
availability of comprehensive corpora like the Helsinki corpus of diachronic English
texts and with the increased processing power of personal computers that scholars
looked to corpora for corroboration or refutation of theses concerning the diachronic
development of languages.
The first decade of the 21st century has seen a great expansion in the compilation
of large corpora, often specialising on a particular variety or text type. The potential
for discovering and documenting minute changes, especially in the major European
languages, which are served well by corpora, has increased greatly in the past decade.
New studies of change which have such corpora as their data basis have appeared, for
instance, Leech, Hundt, Mair and Smith (2009).
8.2 Genre variation and stylistics
The range and size of recent corpora has meant that the study of different genres, style
levels, text types, etc. has received much impetus. Register variation is now fairly easy
to quantify (Biber 1988) and questions of style and not least of authorship can be tackled
with much greater confidence than in the days before large corpora were available. The
increasing engagement of colleagues in university departments in the task of compiling
corpora has led to a welcome diversification of text types. These now encompass not
just written texts but dialogues and transcribed colloquial speech as well.

9. Pathways of change
The movement away from the reconstruction of languages after the heyday of IndoEuropean studies in the 19th and early 20th centuries has meant a reduction of interest
in specific instances of language change. The emphasis moved rather to types of change
and in particular what one could term pathways of change. Two particular directions
in this sphere can be recognized, grammaticalization and typology, both of which have
in common that they are concerned with long-term changes.
9.1 Long-term change: Grammaticalization
The essence of grammaticalization is a transition in category from lexical to grammatical
for an element or elements over a fairly long period of time (Heine & Traugott 1992;

Language change

Pagliuca 1994). The label grammaticalization is used by many scholars so that one can
identify different meanings of the term. One is a formal meaning, e.g. the path of
change from a full independent noun, as with OE lc body, form, to a suffix, ModE -ly
as in friendly. This is a change in form starting with an unbound lexical morpheme and
ending with a bound grammatical morpheme. A second meaning of grammaticalisation is found which is more semantic: it refers to a shift in meaning or an additional
use which rests on figurative usage, e.g. the use of go as a marker of the future in
English or werden become in German for the same purpose. These are cases of metaphorical extension with the original full lexical meaning still available.
The merit of grammaticalization is to have shown that there are specific pathways of change, many of which are largely independent of specific languages. However,
the matter is often more complex because the specific course of grammaticalization
depends on the synchronic state of a language when the process begins and on other
changes which might be running concurrently in a language. This point is made
clearly in an investigation of infinitival to in English, German and Dutch, all Germanic
languages but typologically somewhat different (Fischer 1997).
An example of largely language-independent grammaticalization is that given
in Schwenter & Traugott where they discuss the pathway PERSON > OBJECT >
ACTIVITY > SPACE > TIME > QUALITY with each category on the right more
abstract than that on the left (Schwenter & Traugott 1998:262). Another instance of a
pathway would be the development of reflexive pronouns from intensifiers. This is well
attested in the history of English where in Old English the form sylf/seolf was an intensifier and used much as Modern German selbst. The frequent co-occurrence of these
intensifiers with personal pronouns, as in Christ sealde hine selfne for us Christ gave
himself for us, meant that in time, through coalescence and univerbation (pronoun +
intensifier), reflexive forms developed: him + self himself.
A feature of such pathways, which is stressed by the scholars who formulate them,
is their unidirectionality, i.e. increasingly abstract interpretations of linguistic structures or the shift from pragmatic markers (intensifiers) to grammatical elements tend
to be one-way movements in the history of a language. This view is not uncontested,
see the discussion in Section2.8 above.
9.2 Large-scale changes: The typological perspective
In general one can say that instances of grammaticalization are punctual and do not
necessarily lead to typological re-alignment in a language. But there are developments
which have this effect. For re-alignment to take place, change must be on a broad front.
This stands to reason: for a different organisational principle to arise in a language
there must be sufficient movement across the majority of the languages structures.
Such movement can be seen in phonetic attrition, which affects all inputs irrespective

193

194 Raymond Hickey

of class or category. It can also be seen where an organisational principle is borrowed

through language contact as with shifts in word order, often with a switch in the order
of elements from post-specification to pre-specification or vice versa.
Linguistic typology has led to classifications of languages according to dominant
principles of organisation. Traditionally, there are four types, synthetic (polyfunctional
inflections, flexible word order as in Latin), analytic (few inflections, higher functionalization of word order as in modern English), agglutinating (monofunctional, concatenating inflections as in Finnish or Turkish) and polysynthetic or incorporating (very
high density of opaque inflections as in many native American languages). There are
other axes along which classifications have been suggested. For instance Lehmann
(1992:106) suggests adopting a contentive typology (after Klimov 1977, 1983) in which
the basic distinction among languages is that between types in which government or
agreement is the dominant principle. The government type may be accusative or ergative. The accusative type is the normal Indo-European type, e.g. English or German,
with a separate case for nominative and accusative. The ergative type has a case for the
subject of transitive verbs; the subject of intransitive and the object of transitive verbs
is in a separate case called the absolutive (the parallel in English would be: She saw
her and Her went). In agreement-type languages there is a formal parallelism between
word classes depending on their semantics, e.g. animate nouns and active verbs pair
as would inanimate nouns and stative verbs (Lehmann 1992). Another classification
for this type of language is active where both nouns and verbs fall into an active or
inactive class. Active languages may have two words for concepts rendered by one in
government-based languages, e.g. water (moving or still).
Mention should also be made of the classification of languages according to headmarking or dependent-marking, something which has been the object of detailed
investigation by Johanna Nichols (see Nichols 1992). What is meant here is that in
constructions which involve a head and a dependent, e.g. English the mans house,
where house is the head (the modified element) and man the dependent (the modifier),
those languages which mark the dependent element do so consistently and are labelled
dependent-marking. Other languages, like Hungarian, cf. az ember hza the mans
house, where haz house is marked with the pronominal suffix -a (Fox 1995:266f.), are
termed head-marking. Nichols notes that such marking is very stable over time and
can thus be used as a criterion for the relatedness or non-relatedness of languages.
10. Contact accounts
In accounts of language change, contact looms large. Analyses based on language contact have repeatedly been made because in so many instances contact is a possibility
(Hickey 2010). Furthermore, in some studies, notably Heine & Kuteva (2005), language

Language change

contact scenarios are seen as loci of grammaticalisation processes, an opinion not

shared by all scholars (see Hickey 2007: Chapter 4 for a relevant discussion).
Already in 19th century Indo-European studies contact appears as an explanation
for change though by and large mainstream Indo-Europeanists preferred languageinternal accounts. One should stress that strictly speaking contact is not so much an
explanation for language change as a suggestion for the source of a change, that is, it
does not say why a change took place but rather where it came from. For instance, a
language such as Irish or Welsh may have VSO as a result of early contact with languages also showing this word order. However, this does not explain how VSO arose
in the first place (assuming that it is not an original word order for any language).
The upshot of this is that contact accounts frequently just push back the quest for
explanation a stage further.
Considerable criticism has been levelled at contact accounts because scholars have
often been all too ready to accept contact as a source, to the neglect of internal factors
or inherited features within a language. This readiness to accept contact, particularly
when other possibilities have not been given due consideration, has led to much criticism of contact accounts in the 1970s and 1980s (see Lass 1997 as a representative of
this line of argument). However, a certain swing around can be seen from the 1990s
onwards, a re-valorisation of language contact when considered from an objective and
linguistically acceptable point of view as demanded by Thomason & Kaufman (1988),
see the various contributions in Hickey (ed., 2010).
A further issue in contact studies is just what can be traced to contact. The view
is prevalent that any feature of a language can be the result of contact if the conditions
of contact have been intense and prolonged enough (Thomason & Kaufman 1988;
Thomason 2001). Difficulties arise in situations where an attested feature or structure
can have arisen through contact or retention from earlier input. This issue is a major
one with varieties of English, particularly in the study of overseas forms of English
(Filppula 2002; Hickey ed. 2002).
One should note that language contact is a convenient shorthand for what is, of
course, speaker contact. This fact is not trivial because the results of contact depend
on the nature of speaker contact. In the history of English there is a considerable difference between the contact between Scandinavian and Old English, which was of
an everyday nature, and that between Middle English and Anglo-Norman which was
confined to certain strata of society (the upper clergy, military, aristocracy and court)
not to mention the indirect contact with Central French in the second half of the
Middle English period after the 13th century.
Language contact provides evidence for the autonomy of linguistic levels. In general
it is the open classes, words and phrases, which are most easily affected and hence
borrowings take place here with only slight contact or with what has been termed
cultural contact, indirect contact of the type just mentioned between Middle English

195

196 Raymond Hickey

and Central French. Here there is little or no speaker contact. In contemporary societies
this contact can be seen in the massive influence of English on other languages. It is
more the result of a one-way exposure to English through the media than of any contact in the sense of face-to-face interaction. In such instances, the closed classes of a
language, the phonology, morphology and syntax remain unaffected by this exposure
to another language. Possible influence on closed classes does, of course, depend on
the intensity of the contact and the age of the speakers in contact as well as on the typological distance between the languages involved in a contact situation. For instance,
Swahili (a Bantu language spoken in large parts of eastern Africa) has been influenced
lexically by Arabic for centuries but it has not adopted the stem-inflecting principles
of the latter language.
Returning to the position of speakers in a contact situation, one can distinguish
between a type of contact in which speakers retain their inherited language and one
where they switch to the language they come in contact with. This distinction has been
captured in the two terms contact-induced change and shift-induced change (a distinction stressed by Sarah Thomason, see Thomason 2001). There are differences in the
type of influence which arise in these two kinds of change. Shift-induced change is
seen where speakers abandon their inherited language (see the contributions in Dorian
1989) and move to a new one, usually over a period of at least a few generations. This
has happened historically in many countries, for instance in Ireland with the shift from
Irish to English and in South Africa with the shift from Bhojpuri to English among
the Indian population in KwaZulu-Natal (Mesthrie 1992). Contact with shift leads to
new varieties of a language arising and here it is often the closed classes the sound
and grammar systems which are affected. This is due to the fact that in language
shift (during adulthood, through a process of unguided second language acquisition)
speakers search in the second language for equivalents to categories which they know
from their first language. Historically, this can be clearly seen in the rise of certain
aspectual categories in Irish English. Indeed a case can be made for speakers taking
afunctional elements in the second language, such as the unstressed, declarative do of
early modern English, and employing them for their own purposes, in the case of Irish
English, to express the habitual aspect (Hickey 1997) as in sentences like She does be
worrying about the children She is always worrying about the children.
The relative social status of speakers in a language shift situation is an important
consideration. This is usually captured by the terms substrate and superstrate for
the language in the socially inferior and superior position respectively. If the shifting
group has high social prestige (not the case with the Irish and Indian populations just
alluded to) then they may transfer their speech habits to the speakers of the language
they are shifting to. This is technically known as imposition (Guy 1990) and it has
been proposed that it happened in many historically attested situations, such as the
Scandinavian period in early medieval Scotland, with Vikings who switched to Gaelic,

Language change

or the late medieval period in Ireland with the Anglo-Normans who later switched
completely to Irish. Imposition may account for the appearance of borrowings from
core vocabulary in a contact situation. If, for instance, the Normans retained French
words in their form of Irish, then the Irish themselves may have picked up these words
from the Irish of the Normans who were the military and aristocratic leaders in Ireland
for a few centuries after their first arrival in the late 12th century. This would help to
account for why the words for child (piste < page) and boy (garsn < garon), for
instance, are from Norman French in Irish.
Finally one should mention that contact is an essential part of the twin processes
of pidginization and creolization, but a consideration of these developments is beyond
the remit of the present article.
11. Language areas (Sprachbnde)
The cases of contact discussed so far have been unidirectional. But contact can go
both ways. Specifically, if languages in a contact situation are roughly equal in status
(adstrate languages) then the influence can be mutual. There must be something like
parity between the languages involved otherwise if one gains the upper hand, sociolinguistically speaking, it is unlikely to accept elements from the other language in
the contact situation anymore. If the contact lasts for several centuries then a situation of convergence may arise where certain structural features come to be shared
between languages. The duration of contact is important here: convergence presupposes a long period of stable bilingualism/multilingualism throughout an entire area
in which the languages in question are spoken. This leads to a situation where, to put
it in Lehistes words, genetic heterogeneity is gradually replaced by typological homogeneity (1988:59).
Convergence due to prolonged contact is characteristic of geographically delimited
regions and hence the term linguistic area is often used (Campbell 1998: 299310).
The original German term, however, is Sprachbund language federation which stresses
the common ties between languages and not their geographical proximity. This latter
aspect is probably a necessary but by no means a sufficient condition for the rise of
a Sprachbund.
In the literature on linguistic areas a continuous theme is the number of defining
features necessary. This issue is difficult to resolve to the satisfaction of all scholars in
the field. For instance, is it necessary for all languages in a linguistic area to share all
features used to define the area? This maximal requirement is obviously too strict a
yardstick because even in cases where there is much agreement, such as the Balkans, all
features are not found in all languages. Another issue is the approach to areal features.
Some studies just list features and imply that these have arisen through diffusion,

197

198 Raymond Hickey

while others attempt to trace common features and to account for their appearance by
establishing historical connections and by considering genetic relationships in an area
or the typological unusualness of features (Campbell loc. cit.).
There have been many suggestions for linguistic areas in the relevant literature.
The best known cases are the Balkans, the Indian subcontinent, Mesoamerica, the Baltic
region, the north-west coast of North America and Ethiopia. Other cases, such as the
British Isles (Wagner 1959), or northern Britain and Scandinavia (Borgstrm 1974),
have been considered in literature dating back to the mid 20th century and have been
re-assessed again in the light of recent developments in the field (Hickey 1999).
12. Conclusion
Research into language change is becoming increasingly diversified. There are many
new approaches which seek to expand and extend models already available. Linguists
who formerly might just have looked at a single type of language change have come
to consider several approaches, e.g. principles and parameters along with grammaticalization (Fischer 2006). The historical dimension which has been added to sociolinguistics since the seminal work of Romaine (1982) is a good instance of the expansion
of an established approach. Another approach which has been beneficial to the field
is that of diachronic corpus analysis which has led to many new insights, particularly
into the history of English, and has caused revisions of standard wisdoms. The more
such approaches come together the greater the opportunity for cross-fertilization and
the greater the likelihood that hitherto unobserved shifts in languages can be described
in terms which are factually accurate and linguistically acceptable.

References
Aitchison, J. (2001). Language change: Progress or decay? 3rd ed. Cambridge University Press.
Allwood, J. & P. Grdenfors (eds.) (1999). Cognitive semantics. Meaning and cognition. John Benjamins.
Andersen, H. (1973). Abductive and deductive change. Language 49: 765793.
(1988). Center and periphery: Adoption, diffusion, and spread. In J. Fisiak (ed.) Historical
Dialectology: 3983. Mouton de Gruyter.
Aarts, Bas, David Denison, Evelien Keizer & Gergana Popova (eds) 2004. Fuzzy Grammar A Reader.
Oxford: University Press.
Baldi, P. (ed.) (1991). Patterns of change. Change of patterns. Linguistic change and reconstruction
methodology. Mouton de Gruyter.
Barcelona, A. (ed.) (2000). Metaphor and metonymy at the crossroads. A cognitive perspective.
Mouton de Gruyter.
Bauer, L. (1994). Watching English change. An introduction to the study of linguistic change in standard
Englishes in the twentieth century. Longman.

Language change 199

Beard, A. 2004. Language change. London: Routledge.

Biber, D. (1988). Variation across speech and writing. Cambridge University Press.
Borgstrm, C.H. (1974). On the influence of Norse on Scottish Gaelic. Preaspiration of stops and
pitch patterns. Lochlann 6: 91103.
Brinton, L. & E. Traugott (2005). Lexicalization and language change. Cambridge: University Press.
Britain, D. (1997). Dialect Contact and Phonological Reallocation: Canadian Raising in the English
Fens. Language in Society 26: 1546.
Bynon, R.T. (1977). Historical linguistics. Cambridge University Press.
Campbell, L. (1997). American Indian languages: the historical linguistics of Native America. Oxford:
Oxford University Press.
Campbell, L. (2004 [1998]). Historical linguistics. An introduction. Second edition. Edinburgh
University Press.
Campbell, L. & William J. Poser 2008. Language classification: history and method. Cambridge:
Cambridge University Press.
Chambers, J.K., P. Trudgill & N. Schilling-Estes (eds) (2002). The handbook of language variation and
change. Oxford: Blackwell.
Chen, M. & W. Wang (1975). Sound change: actuation and implementation. Language 51: 255281.
Cheshire, J. (1996). Syntactic variation and the concept of prominence. In J. Klemola, M. Kyt &
M.Rissanen (eds.) Speech past and present: 117. Peter Lang.
Clark, R. & I. Robert (1993). A computational model of language learning and language change.
Linguistic Inquiry 24: 299345.
Crisma, P. & G. Longobardi (2009). Historical Syntax and Linguistic Theory. Oxford: University Press.
Croft, W. (2001). Explaining Language Change. An Evolutionary Approach. London: Longman.
Crowley, T. (1998). An introduction to historical linguistics. 2nd ed. Oxford University Press.
Dasher, R. & E.C. Traugott (2002). Regularity in semantic change. Cambridge University Press.
Denison, D. (1999). Gradience and linguistic change. In L. Brinton (ed.) Historical Linguistics
1999:119144. John Benjamins.
(2003). Log(ist)ic and simplistic S-curves. In R. Hickey (ed.): 5470.
Denison, D. 2004. Do grammars change when they leak?, in Kay, Horobin & Smith (eds),
pp.1530.
Dirven, R. & M. Verspoor (1999). Cognitive exploration of language and linguistics. John Benjamins.
Dorian, N.C. (1993). Internally and externally motivated change in contact situations: doubts about
dichotomy. In Charles Jones, ed., Historical linguistics: problems and perspectives. London:
Longman, pp. 131155.
Dorian, N. (ed.) (1989). Investigating obsolescence. Studies in language contraction and death.
Cambridge University Press.
Durie, M. & M. Ross (eds.) (1996). The comparative method reviewed. Regularity and irregularity in
language change. Oxford University Press.
Fennell, B. (2001). A history of English. A sociolinguistic approach. Blackwell.
Filppula, M. (2002). The quest for the most parsimonious explanations: endogeny vs. contact
revisited. In Hickey (ed.): 161173.
Fischer, O. (1997). The grammaticalisation of infinitival to in English compared with German and
Dutch. In R. Hickey & S. Puppel (eds.): 265280.
(2000). Grammaticalisation: unidirectional, non-reversable? The case of to before the infinitive in English. In O. Fischer et al. (eds.): 137.
(2006). Morphosyntactic Change. Functional and Formal Perspectives. Oxford: University Press.
Fischer, O. & A. Rosenbach (2000). Introduction. In O. Fischer et al. (eds.): 149169.

200 Raymond Hickey

Fischer, O., A. Rosenbach & D. Stein (eds.) (2000). Pathways of Change. Grammaticalization in English.
John Benjamins.
Fischer, O., A. Van Kemenade, W. Koopman & W. van der Wurff (2000). The syntax of early English.
Cambridge University Press.
Fischer, O. & M. Nnny (eds.) (2001). Iconicity. Special issue of the European Journal of English Studies
(EJES) 5(1).
Fox, A. (1995). Linguistic reconstruction. An introduction to theory and method. Oxford University
Press.
Gerritsen, M. & D. Stein (eds.) (1992). Internal and external factors in syntactic change. Mouton
de Gruyter.
Givn, T. 1979. On understanding grammar. New York: Academic Press.
Guy, G.R. (1990). The sociolinguistic types of language change. Diachronica 7: 4767.
Haiman, J. (1983). Iconic and economic motivation. Language 59: 781819.
Haiman, J. (ed.) (1985). Iconicity in syntax. John Benjamins.
Harris, A. & L. Campbell (1995). Historical syntax in cross-linguistic perspective. Cambridge
University Press.
Heine, B. (1997). Cognitive foundations of grammar. Oxford University Press.
Heine, B. & T. Kuteva 2005. Language Contact and Grammatical Change. Cambridge: Cambridge
University Press.
Heine, B. & T. Kuteva 2007. The Genesis of Grammar. A Reconstruction. Oxford: University Press.
Heine, B. & E. Traugott (eds.) (1992). Aspects of grammaticalization. John Benjamins.
Hickey, R. (1997). Arguments for creolisation in Irish English. In R. Hickey & S. Puppel (eds.):
9691038.
(1999). Ireland as a linguistic area. In J.P. Mallory (ed.) Language in Ulster. Special issue of
Ulster Folklife (45): 3653.
(2002c). Ebb and flow. A cautionary tale of language change. In T. Fanego, B. Mendez-Naya &
E. Seoane (eds.) Sounds, words, texts, change: 105128. John Benjamins.
(2003a). Reanalysis and typological change. In R. Hickey (ed.): 258278.
(2003b). How do dialects get the features they have? On the process of new dialect formation.
In R. Hickey (ed.): 213239.
(2003c). Rectifying a standard deficiency. Pronominal distinctions in varieties of English.
In A.H. Jucker & I. Taavitsainen (eds.) Diachronic perspectives on address term systems. John
Benjamins.
(2007). Irish English. History and Present-day Forms. Cambridge: Cambridge University Press.
(2010). Language Contact: Reassessment and reconsideration, in: Raymond Hickey (ed.)
The Handbook of Language Contact. Malden, MA: Wiley-Blackwell, pp. 128.
Hickey, R. (ed.) (2003). Motives for language change. Cambridge University Press.
Hickey, R. (ed.) (2010). The Handbook of Language Contact. Malden, MA: Wiley-Blackwell.
Hickey, R. & S. Puppel (eds.) (1997). Language history and linguistic modelling. A festschrift for Jacek
Fisiak on his 60th birthday. Mouton de Gruyter.
Ihalainen, O. (1994). The dialects of England since 1776. In R. Burchfield (ed.) The Cambridge history
of the English language. Vol. 5: 197274. Cambridge University Press.
Janda, R.D. (2001). Beyond pathways and unidirectionality: On the discontinuity of language
transmission and the counterability of grammaticalization. Language Sciences 23(23): 265340.
Jansen, T. & G. Redeker (eds.) (2000). Cognitive linguistics. Foundations, scope, methodology. Mouton
de Gruyter.

Language change

Jones, M. & I. Singh (2005). Exploring language change. London: Routledge.

Joseph, B. (2001). Is there such a thing as grammaticalization? Language Sciences 23(23): 163186.
Jucker, A. (ed.) (1998). Historical pragmatics. John Benjamins.
Van Kemenade, A. & N. Vincent (eds.) (1997). Parameters of morphosyntactic change. Cambridge
University Press.
Kenstowicz, M. & C. Kisseberth (1979). Generative phonology. Description and theory. Academic
Press.
Kim, H. (2001). Remarks on the unidirectionality principle in grammaticalization. Folia Linguistica
Historica 22(12): 4966.
King, R. (1969). Historical linguistics and generative grammar. Prentice Hall.
Kittay, E. (1987). Metaphor. Its cognitive force and linguistic structure. Oxford University Press.
Klimov, G.A. (1977). Tipologija jazykov aktivnogo stroja. [A typology of languages of the active kind.]
Nauka.
(1983). Principy kontensivnoj tipologi. [The principles of contentive typology.] Nauka.
Koutsoudas, A. (ed.) (1976). The application and ordering of grammatical rules. Mouton.
Kroch, A. (1989). Reflexes of grammar in patterns of language change. Journal of Language Variation
and Change 1: 199244.
Labov, W. (1977). On the use of the present to explain the past. In: Adam Makkai, Valerie Becker
Makkai & Luigi Heilmann (eds) Linguistics at the Crossroads. Lake Bluff, IL: Jupiter Press,
pp.226261.
(1981). Resolving the Neogrammarian controversy. Language 57: 267308.
(1994). Principles of linguistic change. Vol. 1: Internal factors. Blackwell.
(2001). Principles of linguistic change. Vol. 2: Social factors. Blackwell.
(2010). Principles of linguistic change. Vol. 3 Cognitive and cultural factors. Blackwell.
Lahiri, A. (ed.) (2000). Analogy, levelling, markedness. Principles of change in phonology and morphology.
Mouton de Gruyter.
Langacker, R.W. (1991). Concept, image and symbol. The cognitive basis of grammar. Mouton de Gruyter.
(2000). Grammar and conceptualization. Mouton de Gruyter.
Lappin, S. (1996). The handbook of contemporary semantic theory. Blackwell.
Lass, R. (1980). On explaining language change. Cambridge University Press.
(1997). Historical linguistics and language change. Cambridge University Press.
(2001). Remarks on (uni)directionality. In O. Fischer et al. (eds.): 207227.
Leech, G., M. Hundt, C. Mair & N. Smith 2009. Change in Contemporary English: A Grammatical
Study. Cambridge: University Press.
Lehiste, I. (1988). Lectures on language contact. MIT Press.
Lehmann, W.P. (1992). Historical linguistics. 3rd ed. Routledge.
Lightfoot, D. (1979). Principles of diachronic syntax. Cambridge University Press.
(1991). How to set parameters: Arguments from language change. MIT Press.
(1999). The development of language. Blackwell.
(2002). Grammaticalization: Cause or effect? In R. Hickey (ed.): 99123.
Mac Cormac, E.R. (1985). A cognitive theory of metaphor. MIT Press.
McMahon, A. (1994). Understanding language change. Cambridge University Press.
Mesthrie, R. (1992). English in language shift. The history, structure and sociolinguistics of South African
Indian English. Cambridge University Press.
Milroy, J. (1992). Linguistic variation and change. Blackwell.
(2002). On the role of the speaker in language change. In R. Hickey (ed.): 143157.

201

202 Raymond Hickey

Milroy, J. & J. Harris (1980). When is a merger not a merger? The MEAT/MATE problem in a presentday English vernacular. English World-Wide 1: 199210.
Newmeyer, F.J. (1998). Language form and language function. MIT Press.
(2002). Formal and functional motivation for language change. In R. Hickey (ed.): 1836.
Nichols, J. (1992). Language diversity through space and time. Chicago University Press.
Norde, M. 2009. Degrammaticalization. Oxford: University Press.
Orton, H. & W.J. Halliday (1963). Survey of English Dialects. Arnold.
Pagliuca, W. (1994). Perspectives on grammaticalization. John Benjamins.
Paul, H. (1975 [1888]). Prinzipien der Sprachgeschichte. [The principles of language history.] Niemeyer.
Phillips, B. (2006). Word Frequency and Lexical Diffusion. London: Palgrave Macmillan.
Pintzuk, S. (1991). Phrase Structures in Competition: Variation and Change in Old English Word Order.
Ph.D. Diss. University of Pennsylvania.
Pintzuk, S., G. Tsoulas & A. Warner (eds.) (2001). Diachronic syntax. Models and mechanisms. Oxford
University Press.
Roberts, I. (1985). Agreement parameters and the development of English modal auxiliaries. Natural
Language and Linguistic Theory 3: 2158.
(1993). Verbs and Diachronic Syntax. Kluwer.
Roberts, I. (2007). Diachronic Syntax. Oxford: University Press.
Romaine, S. (1982). Socio-Historical Linguistics: its Status and Methodology. Cambridge University
Press.
Schwenter, S.A. & E.C. Traugott (1998). The semantic and pragmatic development of substitutive
complex prepositions in English. In A. Jucker (ed.): 243274.
Thomason, S.G. (2001). Language contact: An introduction. Edinburgh University Press.
(2003). Contact as a source of language change, in: Joseph, Brian D. & Richard D. Janda (eds)
The Handbook of Historical Linguistics. Oxford: Blackwell, pp. 687712.
Thomason, S.G. & T. Kaufman (1988). Language contact, creolization, and genetic linguistics. University
of California Press.
Trask, R.L. (1996). Historical linguistics. Arnold.
Trask, R.L. & R. McColl Millar (2010). Why Do Languages Change? Cambridge: Cambridge University
Press.
Traugott, E.C. (1989). On the rise of epistemic meanings in English: An example of subjectification
in semantic change. Language 57: 3365.
Traugott, E. & R. Dasher (2002). Regularity in semantic change. Cambridge: Cambridge University
Press.
Trudgill, P. (1986). Dialects in contact. Blackwell.
(1999). A window on the past: colonial lag and New Zealand evidence for the phonology of
19th-century English. American Speech 74(3): 111.
Ungerer, F. & H.-J. Schmid (1996). An introduction to cognitive linguistics. Longman.
van der Auwera, J. (2002). More thoughts on degrammaticalization. In I. Wischer & G. Diewald
(eds.): 1930.
Wagner, H. (1959). Das Verbum in den Sprachen der britischen Inseln. Niemeyer.
Wang, W. (1969). Competing changes as a cause of residue. Language 45: 925.
Weinreich, U., W. Labov & M. Herzog (1968). Empirical foundations for a theory of language
change. In W. Lehmann & Y. Malkiel (eds.) Directions for historical linguistics: 95195. University
of Texas Press.
Wierzbicka, A. (1996). Semantics. Primes and universals. Oxford University Press.
Wischer, I. & G. Diewald (eds.) (2002). New Reflections on Grammaticalization. John Benjamins.

Language contact
Yaron Matras

1. Introduction
Language contact is the interplay of two or more linguistic systems. Research into
language contact examines the potential consequences of this interplay in usage patterns
as well as in the structure of language. Since contact involves factors that are external
to the linguistic system (in the canonical understanding of system), language contact
had initially been associated primarily with sociolinguistic research agendas. Since the
late 1970s, however, there has been growing interest in language contact also within
psycholinguistics, historical linguistics, and linguistic typology. Societal and individual aspects of language contact those having to do with choices that speakers make
among linguistic repertoires are often discussed under the headings bilingualism
or multilingualism.
Most influential in establishing language contact and bilingualism as a subject of
study was Uriel Weinreichs Languages in contact (Weinreich 1953). Work in the field
has since taken various specialized paths, with focus areas devoted to societal aspects
(language use and language shift, language planning), individual aspects (the bilingual
brain, parallel acquisition of two languages), conversational aspects (language alternation and language mixing in conversation), grammatical aspects (structural constraints
on language mixing, contact-induced structural change), and typological-classificational
aspects (the emergence and structure of contact languages and linguistic areas).
2. Societal multilingualism
Fishman (1967) introduced a classification grid for multilingual societies which recognized that widespread bilingualism is usually accompanied by diglossia, i.e. a functional specialization of the languages. Language use in a multilingual community can
be surveyed and mapped by domains, thus addressing the question who speaks
what language to whom and when? (Fishman 1965). Stable bilingualism without
diglossia is regarded by Fishman (1967) as a typically transitional stage. It may be
found for instance among the second generation of immigrants, who are integrated
linguistically, having been raised in the new country while still retaining full competence of the language of their country of origin. This instability of bilingualism
thus requires continuous re-negotiation of the social roles played by the individual

204 Yaron Matras

languages involved. From a somewhat more extreme viewpoint, it has even been
claimed, therefore, that there is seldom language contact without language conflict
(Nelde 1980), as speakers of smaller languages struggle for a balance between linguistic separateness as a token of identity and the socio-economic benefits of proficiency
in the language of a more powerful group, or attempt to fend off political pressure to
abandon their language. Language abandonment itself language shift is perceived
of as a gradual retreat in the use of a language by individual speakers and in individual
domains (Gal 1979), leading in some cases to the near disappearance of a language
(Dorian 1981). During the late 1990s, interest in language death and language endangerment intensified through the realization that much of the worlds linguistic diversity
is threatened (e.g. Crystal 2000).
3. The bilingual individual
The question of the mental storage and activation of two languages has been in the
center of research in bilingualism since its early stages. Weinreich (1953) introduced
a distinction between compound bilinguals, who have two inventories of signs at
their disposal representing a single conceptual lexicon, and coordinate bilinguals,
whose languages are distinguished also at the conceptual level. The distinction was
later understood as reflecting different contexts of acquisition compound relating
to acquisition of two languages in the same cultural context, coordinate in separate
contexts (Ervin & Osgood 1954). The extent to which such a distinction can be supported empirically has remained controversial. Experimental studies of speakers
reactions to stimuli lend some support to a correlation between context of acquisition
and use, and accessibility in language processing. Bilingual aphasics have also been
reported to show different recovery patterns for their individual languages, though no
general correlation can be found between context of acquisition and recovery pattern.
There is some evidence to suggest that language processing involves stronger righthemispheric activities among bilinguals than among monolinguals, but this could
be explained by the prominence of pragmatic-situative factors that are involved in
acquiring and communicating in a second language, rather than by cerebral separation of the languages themselves (Albert & Obler 1978). The consensus view has since
been formulated by Paradis (1987) as neurofunctionally independent, but not neuroanatomically separated.
Another context in which the degree of autonomy of the two systems has been
discussed is the parallel acquisition of two languages, or Bilingual First Language
Acquisition. Advocates of the Unitary Language System Hypothesis have argued that
during the initial stages of acquisition, the child is unable to distinguish between the
systems and mixes them indiscriminately. Volterra & Teaschner (1978) proposed a

Language contact 205

three-stage model. At first the child has one combined lexicon. This is followed by
gradual differentiation of the lexicon but just one syntactic system, and finally by the
beginning differentiation of the languages in grammar as well. Age correlates of these
phases are highly individual and so generalizations are difficult to make. The Unitary
Language System Hypothesis has come under criticism, however. Its challengers have
pointed out that overgeneralizations and use of pragmatic principles of sentence organization (rather than fully acquired syntactic rules) among young children can easily
bias the picture, giving the impression that grammatical structures are mixed. Meisel
(1989) cited evidence for early syntactic differentiation among bilingual children,
and argued for a separation between mixing as part of the grammatical competence
(i.e. inability to differentiate two system), and code-switching as part of a bilinguals
pragmatic competence. This approach has been influential in determining the stateof-the-art in bilingual language acquisition research. The current trend, best represented by Lanzas (1997) work, regards bilingual language development as the gradual
acquisition of sociolinguistic and pragmatic rules governing the appropriateness of
linguistic choices. Lanza was able to illustrate a direct correlation between adults
response strategies to wrong or inappropriate language choices, and the childs ability
to sustain differentiation of the systems in conversation.
4. Language mixing in conversation
The established terms for adult language mixing are code-mixing or code-switching;
the latter is used by some authors, though not by all, strictly to denote intersentential
mixing or mixing between utterances. The distinction is also captured by the terms
insertion, which refers to the employment of lexical material from L2 in an L1 sentence
framework, and (code) alternation, which implies switching at the utterance boundary.
A further term, congruent lexicalization, has been used to denote the occurrence of
from two or more languages in a sentences structure that is shared, fully or in part
(Muysken 2000).
Speech situation and topic are important factors influencing the choices of code
that speakers make. It is accepted that speakers choices may involve complex strategies
of accommodation to a variety of factors, including the identity and relative prestige
of the interlocutor as well as the setting and topic (Gardner-Chloros 1991). Moreover, code-switching is indexical of negotiating extra-linguistic relations among
participants, as speakers are guided by a consensus on the relative markedness of a
language choice in a specific exchange (Myers-Scotton 1993a). In addition to extralinguistic factors such as setting, topic, and interlocutor, language-internal factors may
also serve as triggers (Clyne 1967) for language choices. Classic linguistic triggers are
vocabulary gaps, as well as structural similarities between the languages. Conversation

206 Yaron Matras

analysis has recognized conversation-internal, pragmatic motivations for switching.

According to Gumperz (1982), switches can act as contextualization cues, i.e. they
have a signaling function, marking out boundaries of units in the discourse. A typical
metaphorical switch is one that is carried out in order to authenticate a quotation in
narrations (by quoting in the language of the original utterance). Apart from reported
speech, typical conversational loci of switching are parentheses or side-comments,
reiterations, puns or language play, change of activity type, or topic shift. Auer (1984)
emphasizes the importance of sequentiality of language choice across turns and distinguishes discourse-related code-switching (which contributes to the organization of
the discourse) from participant-related code-switching (which involves expressing or
negotiating language preferences). Alongside the investigation of conversational motivations for code-switching there is also recognition that for some groups of speakers
or even for entire communities, switching itself might be the default or unmarked
choice (Meeuwis & Blommaert 1998).
A central issue in code-switching research has been the attempt to formulate
structural regularities. Poplack (1980) proposed the Equivalence Constraint, according
to which code-switches will tend to occur at points where the juxtaposition of the
two languages does not violate the syntactic rules of either language, and the Free
Morpheme Constraint, according to which switches cannot occur after bound morphemes. Violations of the Free Morpheme Constraint were defined in subsequent
work (e.g. Poplack, Sankoff & Miller 1988) as nonce-borrowings, i.e. instances of full
integration of an element into the base language. A constraint based on Government &
Binding theory was proposed by DiSciullo, Muysken & Singh (1986). They predicted
the ungrammaticality of mixes between elements that are bound by a government or
selection relation, such as verb/object, preposition/NP-complement, or verb/clausal
complement, while on the other hand ungoverned elements, such as tags, exclamations, and most adverbs are frequent switching points. A model that relates language
mixing to categorial structure in the lexicon is proposed by Myers-Scotton (1993b).
Her Matrix Language Frame model assumes an asymmetrical relationship between the
Embedded Language (EL), which supplies Content Morphs, and the Matrix Language
(ML), which supplies the grammatical frame of the sentence (System Morphs).
Muysken (2000) distances himself from absolute constraints, due to the difficulty
of accommodating the large number of counter-examples, and favors instead those
predictions that are probabilistic, foreseeing general trends. Statistics on the frequency
of switched categories may be of relevance here, indicating a general trend for nouns
to be switched more often than verbs, which in turn are switched more often than
adjectives, which are then followed by other categories. Variation in the hierarchies
may depend on typological factors (relating to individual pairs of languages) as well
as on the switching habits of particular communities. Interestingly, directionality may
play an important role. Thus in Poplacks (1980) study of Spanish-English bilingualism

Language contact 207

in New York, the amount of English-to-Spanish switching around tags and interjections was negligible, while among the same speakers Spanish-to-English switches of
the very same elements were extremely frequent. Poplack interpreted these findings
as reflecting emblematic code-switching by speakers who are less proficient in English,
but seek to display their bilingual competence nonetheless, thus emphasizing those
English items that are more easily acquired and more easily integrated. There has since
been growing interest in discourse-functional explanations, especially for the occurrence of bilingual discourse markers. Salmons (1990) described the adoption of discourse markers as a process of convergence of communication strategies. Maschler
(1994) considered the insertion of L2 discourse markers as a metalinguistic conversational technique to highlight boundaries between units of discourse, while Matras
(1998) has suggested that their gesture-like function makes discourse markers prone
to bilingual speech production errors and susceptible to a takeover through the pragmatically dominant language.
Many constraints on code-switching have been formulated on the basis of the
occurrence of switches in sentences that observe the rules on grammatical wellformedness of both languages. But switching behavior has also been shown to lead
to changes in what speakers consider acceptable or grammatical. Structural convergence in the speech of bilinguals may involve the combined patterns of more than
one language, or use of material from one language modeled on the syntactic rules of
another. The common explanation for such developments is the ease of processing,
which promotes uniformity both at the level of sentence arrangement, and possibly
also at the level of the functional scope covered by individual categories, such as tense
and aspect categories (Silva-Corvalan 1994). Such cases of emergent bilingual grammar
can be regarded as a Composite Matrix Language (Bolonyai 1998; Jake 1998).
5. Contact-induced language change
Language contact can result in changes through the impact of one system on another.
This is expressed by the earlier terms interference or (negative) transfer, which highlighted the possible effect of bilingualism on an individuals performance, while the
term borrowing (and less frequently loans), although repeatedly criticized in the
literature, remains the most common term to denote the replication of a structure
from a source or donor language in another, recipient language (e.g. Haugen 1950;
Lehiste 1988). The most widely discussed borrowing in traditional historical linguistic
literature (not specialized in contact) was lexical, in particular so-called cultural loans,
adopted in conjunction with new concepts or artifacts. This attention to cultural loan
vocabulary biased somewhat the early discussion of grammatical borrowing, which
considered gaps in the recipient language to be the triggers of the process. Indeed, even

208 Yaron Matras

in recent literature, gaps are not ruled out as a motivation for grammatical borrowing
(e.g. Harris & Campbell 1993).
An issue of continuous debate in the literature is the extent to which general principles or constraints on borrowing may be formulated. There are however not just
different viewpoints, but it seems also different understandings of what the notion of a
constraint entails. Several authors have argued against any restrictions on borrowing,
in the sense that there is no structural category that is entirely immune to borrowing
(Campbell 1993; Thomason 2001). There are nonetheless some tendencies and regularities of borrowing that cannot be ignored, even if counter-examples may often be
found. Some of the universals proposed by Moravcsik (1978) suggest for example that
referential autonomy of the structure is a factor promoting borrowability: lexical items
are borrowed earlier than non-lexical items, nouns are more easily borrowed than
non-nouns, free morphemes are more easily borrowed than bound morphemes, and
derivational elements more easily than inflectional elements. Moravcsiks statements
are formulated in part in the form of implication hierarchies; for instance, no inflectional elements will be borrowed unless derivational elements are also borrowed.
Thomason & Kaufmans (1988) borrowability scale hypothesizes that there is a
correlation between duration and intensity of cultural contact, and the identity of
structural categories that are borrowed. Although Thomason & Kaufman do not
recognize absolute constraints on borrowing, they admit that individual categories
vary in their susceptibility to borrowing. At the top of the hierarchy is the lexical
borrowing of content words (Category 1), which can be brought about through just
casual contact. More intense contact may result in borrowing of function words and
minor phonological, syntactic, and lexical semantic features (Category 2). This is followed by borrowing of adpositions, derivational suffixes, and phonemes (Category3).
Strong structural pressure may lead to borrowing of word order patterns, distinctive features in phonology, and inflectional morphology (Category 4). Finally, very
strong cultural pressure may show significant typological disruption and phonetic
changes (Category5).
However, in order to corroborate hierarchical statements of this kind through
quantifiable data, it is necessary to have a sample of various contact situations, involving typologically distinct language pairs, where duration and intensity of contact as
well as the relative sociolinguistic status of the languages involved are comparable or
at least in some way controllable within the sample. The current state-of-the-art in
language contact studies lacks such a comprehensive sample (but see remark below).
Hence, much of the discussion revolves around identifying counter-examples to proposed hierarchies. Ross (1996), for instance, has shown, for languages of Papua New
Guinea, that syntactic-typological convergence may occur without massive borrowing
of lexicon, function words, or morphemes. Ross defines such cases of convergence as
metatypy, which can be taken to mean the adoption or generalization of matching

Language contact 209

patterns of form-function mapping in the languages involved. Thomason (2001) considers as borrowings even the deliberate mixes that lead to the emergence of in-group
mixed languages, which sometimes show wholesale replacement of the basic lexicon,
or adoption of entire pronominal paradigms (see below). It seems, therefore, that the
precise borderline between borrowing in the sense of a gradual adoption and replication of a form through natural communication in language contact situations, and
other related phenomena, is a matter on which consensus is yet to be established.
There is nonetheless tentative acceptance that there is a difference between widespread patterns of borrowings, lexical and grammatical, and what has been referred to
as heavy borrowing. Languages with heavy borrowing show alongside extensive lexical
borrowing also significant influence of the contact language on grammatical categories. Asia Minor Greek as described by Dawkins (1916) adopts Turkish indefinites, the
interrogative marker, comparative and superlative markers, coordinating and subordinating conjunctions, postpositions, adverbs, higher numerals, and the typology of
genitive case marking and relative clause formation. Furthermore, is shows occasional
use of Turkish person inflection on the verb, verb-derivational markers, and uses the
Turkish inflectional marker in the genitive-possessive construction. Chamorro has
borrowed from Spanish a demonstrative, some interrogatives and indefinites, the
comparative marker, all numerals, modality markers, subordinating conjunctions,
prepositions, and sentence adverbs, though no inflectional or derivational morphology.
Domari (the Indo-Aryan language of the Doms or Middle Eastern Gypsies; Matras
1999) has Arabic prepositions, conjunctions and resumptive pronouns, modal and
aspectual auxiliaries (with their inflection), comparative and superlative forms, negators,
numerals above five, interrogatives, phasal adverbs and focus particles.
Languages with heavy borrowing are tentatively so termed since they come
close to showing us the limits of borrowing. This does not mean that it is impossible
to borrow categories such as person and tense inflection on the verb, definite and
indefinite articles, or personal pronouns. Borrowings in these domains are indeed
attested. However, they remain rare in contact situations, even where heavy borrowing is involved, which suggests that grammatical categories do indeed differ in
their universal susceptibility to contact-induced change. Stolz & Stolz (1994) attempt
to formulate a language-internal cline of borrowing susceptibility, according to the
level of syntactic operation. At the top of the susceptibility hierarchy are categories
that operate at the text level, followed by the paragraph level, the sentence and clause
levels, and finally the phrase and word levels. In Matras (1998) it was proposed that
borrowing hierarchies are sensitive to functional properties of discourse organization
and speakerhearer interaction. Items expressing contrast and change are more likely
to be borrowed than items expressing addition and continuation. Discourse markers
such as tags and interjections are on the whole more likely to be borrowed than conjunctions, and categories expressing attitudes to propositions (such as focus particles,

210 Yaron Matras

phasal adverbs like still or already, or modals) are more likely to be borrowed than
categories that are part of the propositional content itself (such as prepositions, or
adverbs of time and place). Contact susceptibility is thus stronger in categories that
convey a stronger link to hearer expectations, indicating that contact-related change
is initiated through the convergence of communication patterns (cf. Salmons 1990).
In the domain of phonology and phonetics, sentence melody, intonation and tones
appear more susceptible to borrowing than segmental features. One might take this
a step further and suggest that contact first affects those functions of language that
are primary or, in evolutionary perspective, primitive. Reacting to external stimuli,
seeking attention, and seeking common ground with a counterpart or interlocutor.
Contact-induced language change thus has the potential to help illuminate the internal
composition of the grammatical apparatus, and indeed even its evolution. One should
like to see these questions pursued through broader sampling and more intense cooperation between descriptive linguists, neurolinguists and psycholinguists.
6. Contact, typology and language classification
Beyond its societal and individual effects, and its effects on the selected structures in
individual languages, language contact may trigger processes of radical typological
re-structuring. Wide attention has been given to pidgins and creoles, the study of
which has, since the 1970s, become a discipline in its own right (cf. Holm 2000). The
earlier phases in the discussion of pidgin and creoles tended to focus on the types of
situations that lead to their emergence, an enumeration of their shared features, the
relations among them, as well as their relevance to a general discussion of language
genesis (on the latter cf. Bickerton 1981). The attention of language typologists has
more recently been drawn to the question of the extent to which creoles actually constitute language types (McWhorter 1998, 2001).
Typologists have also become increasingly interested in the phenomenon of
linguistic areas and the areal clustering of structures. The notion that areal affinity
can be at least as indicative of the structural composition of a language as its genetic
affiliation, and perhaps even more so, emerged already between the 1920s1940s, and
was captured by the terms Sprachbund or linguistic league (Trubetzkoy) and areal
convergence (Weinreich). From the point of view of typological universals, linguistic
areas provide a challenge by showing correlations among structures that are due to
the historical relationship among contiguous speech communities, rather than due to
language-internal constraints on typological harmony. By the traditional definition,
linguistic areas consist of three or more languages that share a set of features that are
not directly inherited. The most widely discussed and best documented linguistic area
is the Balkans, but the discussion of the past decades has identified numerous other

Language contact

linguistic areas around the world. Important contributions to areal linguistics as

the study of linguistic areas is known are Masicas (1976) enumeration of shared
features as an indication of typological proximity within an area, and Josephs (1983)
diachronic study showing how an inherited option can become generalized in an area
through geographical diffusion. In terms of the trigger to convergent development,
Thomason (2001) distinguishes prolonged bilingualism (as in the case of the Balkans)
from imperfect learning of a superstrate language (as in the case of the Ethiopian highlands). But the concept of areal diffusion has received a broader interpretation and
broader relevance in recent years, and it is now increasingly understood as the spread
of isoglosses across linguistic boundaries, some of them clustering in a higher density
in some areas. Thus, the reduction of the infinitive is a famous feature of the Balkans,
which, however, is shared with neighboring regions to the east of the area (cf. Masica
2001). Future tasks of areal linguistics are to determine the extent to which typological
phenomena tend to cluster around the world (cf. Comrie et al. forthcoming), as well as
to identify those categories that may be more susceptible to areal convergence. There
is tentative agreement that these may include tones, definite articles, noun classifiers,
grammaticalized aktionsart modification, word order rules, the typology of modal
complementation, and more. Still to be determined is also the relation between crosslinguistic convergence and conversational convergence strategies in bilingual speech.
Another challenge to typological and historical-linguistic theory are languages
that arise rapidly (within one or two generations) in situations of full bilingualism and
which display mixing patterns that are unconventional. Termed Mixed Languages
in recent literature (Bakker & Mous 1994; Bakker 1997; Thomason 1997; Bakker &
Matras 2003), they tend to defy conventional genetic-classification criteria by showing
a split between the source language for most of the basic vocabulary, and the source
language for much if not most of the grammar. This so-called lexicon:grammar-split
is now increasingly understood as the outcome of deliberate language mixing used
to flag a distinct ethnic identity. In this respect, mixed languages resemble secret or
special languages; indeed, a number of varieties dealt with as mixed in the literature function as in-group registers, used occasionally at the discourse or even just
the utterance level. Mixed languages such as Michif (French-Cree), Mednyj Aleut
(Aleut-Russian), Media Lengua (Spanish-Quechua), or Maa (Cushitic-Pare), display
different kinds of structural-etymological splits. The prototype of a mixed language
appears to consist of a split between the source language that provides finite verb
inflection (or INFL-language; cf. Matras 2004), and the language from which the bulk
of the basic lexicon is derived. This unusual, originally intentional and conscious mixture seems to license further integration of categories that are not normally borrowed
in conventional situations of contact, such as the incorporation of entire pronominal
paradigms (e.g. from Spanish, in a Quechua framework in Media Lengua), or even
of the entire noun phrase grammar (e.g. from French, in a Cree sentence framework

211

212

Yaron Matras

in Michif). The discovery of mixed languages has potential implications for the genetic
classification of isolated languages or problem cases.

7. Concluding remarks
Language contact research continues to address the interface of extralinguistic (social)
circumstances and the development of linguistic structures, and so it demands an
approach that combines sociolinguistic methodology with historical and descriptive
linguistic theory. Thomason (2001) identifies a number of distinct contact mechanisms that take into account the role played by contact in conversation: switching
and alternation, second language and bilingual acquisition, and so on. Terms such as
metatypy (Ross 1996), fusion (Meisel 1989; or in a different sense Matras 1998), or
intertwining (Bakker 1997) show attempts to identify distinct structural outcomes,
relating to specific contact constellations (respectively: multilingualism, bilingual first
language acquisition, bilingual speech production errors, ethnically hybrid communities). The problematic aspect of the combination of sociolinguistic and structural factors is the limitations on the predictive power of generalizations, due to the multiplicity
of factors that enter the equation. Nonetheless, a number of samples of different languages in contact under comparable sociolinguistic and typological conditions have in
recent years supplied background material for a number of generalized observations
(cf.Stolz& Stolz 1994 on Mesoamerica; Bakker & Mous 1994 on mixed languages;
Ross 1996 on Papua New Guinea; Matras 1998 and 2002: Ch. 8 on Romani; Thomason
2001 on linguistic areas). With stronger sampling and a growing number of studies to
compare, language contact is increasingly appreciated as a field that can make a significant contribution to grammatical theory. Myers-Scotton & Jakes (2000) division of
the grammatical system into four distinct levels, based on the behavior of individual
categories in different types of contact, is both constitutive and representative of a
trend within contact studies: The focus is being directed more intensively at language
contact as indicative of the internal structuring of grammar.

References
Albert, M.L. & L.K. Obler (1978). The bilingual brain. Neuropsychological and neurolinguistic aspects
of bilingualism. Academic Press.
Auer, P. (1984). Bilingual conversation. John Benjamins.
Bakker, P. (1997). A language of our own: the genesis of Michif, the mixed Cree-French language.
Oxford University Press.
Bakker, P. & Y. Matras (eds.) (2003). The mixed language debate. Mouton de Gruyter.

Language contact

Bakker, P. & M. Mous (eds.) (1994). Mixed languages. 15 case studies in language intertwining. IFOTT.
Bickerton, D. (1981). Roots of language. Karoma.
Bolonyai, A. (1998). In-between languages: language shift/maintenance in childhood bilingualism.
International Journal of Bilingualism 2: 2143.
Campbell, L. (1993). On proposed universals of grammatical borrowing. In H. Aertsen & R.J. Jeffers
(eds.) Historical linguistics 1989:91109. John Benjamins.
Clyne, M. (1967). Transferance and triggering. Nijhoff.
Comrie, B., M. Dryer, D. Gil & M. Haspelmath (eds.) (forthcoming). World atlas of the linguistic
structures.
Crystal, D. (2000). Language death. Cambridge University Press.
Dawkins, R.M. (1916). Modern Greek in Asia Minor. Cambridge University Press.
Disciullo, A.-M., P. Muysken & R. Singh (1986). Code-mixing and government. Journal of Linguistics
22: 124.
Dorian, N. (1981). Language death. The life cycle of a Scottish Gaelic dialect. University of Pennsylvania
Press.
Ervin, S. & C. Osgood (1954). Second language learning and bilingualism. Journal of Abnormal and
Social Psychology Supplement 49: 139146.
Fishman, J. (1965). Who speaks what language to whom and when. La linguistique 2: 6788.
(1967). Bilingualism with and without diglossia; diglossia with and without bilingualism.
Journal of Social Issues 23: 2938.
Gal, S. (1979). Language shift. Social determinants of linguistic change in bilingual Austria. Academic
Press.
Gardner-Chloros, P. (1991). Language selection and switching in Strasbourg. Clarendon.
Gumperz, J. (1982). Discourse strategies. Cambridge University Press.
Harris, A. & L. Campbell (1995). Historical syntax in cross-linguistic perspective. Cambridge University
Press.
Haugen, E. (1950). The analysis of linguistic borrowing. Language 26: 210231.
Holm, J. (2000). An introduction to pidgins and creoles. Cambridge University Press.
Jake, J. (1998). Constructing interlanguage: building a composite matrix language. Linguistics
36:333382.
Joseph, B. (1983). The synchrony and diachrony of the Balkan infinitive. Cambridge University Press.
Lanza, E. (1997). Language mixing in infant bilingualism: A sociolinguistic perspective. Oxford
University Press.
Lehiste, I. (1988). Lectures on language contact. MIT Press.
Maschler, Y. (1994). Metalanguaging and discourse markers in bilingual conversation. Language in
Society 23: 325366.
Masica, C. (1976). Defining a linguistic area: South Asia. Chicago University Press.
(2001). The definition and significance of linguistic areas. In: P. Bhaskararao & K. Subbarao
(eds.) The Yearbook of South Asian Languages & Linguistics 2001:205268. Sage.
Matras, Y. (1998). Utterance modifiers and universals of grammatical borrowing. Linguistics 36:281331.
(1999). The state of present-day Domari in Jerusalem. Mediterranean Language Review 11:158.
(2003). Romani: A linguistic introduction. Cambridge University Press.
(2004). Re-examining the structural prototype in mixed languages. In P. Bakker & Y.Matras (eds.).
McWhorter, J. (1998). Identifying the Creole prototype: Vindicating a typological class. Language
74: 788818.
(2001). The worlds simplest grammars are creole grammars. Linguistic Typology 5: 125166.

213

214 Yaron Matras

Meeuwis, M. & J. Blommaert (1998). A monolectal view of code-switching. Layered code-switching
among Zairians in Belgium. In P. Auer (ed.) Code-switching in conversation. Languages, interaction
and identity: 7698. Routledge.
Meisel, J.M. (1989). Early differentiation of languages in bilingual children. In K. Hyltenstam &
L.Obler (edss) Bilingualism across the lifespan. Acquisition, maturity and loss: 1340. Cambridge
University Press.
Moravcsik, E. (1978). Universals of language contact. In J. Greenberg (ed.) Universals of human
language: 94122. Stanford University Press.
Muysken, P. (2000). Bilingual speech. Cambridge University Press.
Myers-Scotton, C. (1993a). Social motivations for codeswitching: evidence from Africa. Oxford
University Press.
(1993b). Duelling languages: grammatical structure in codeswitching. Oxford University Press.
Myers-Scotton, C. & J. Jake (2000). Four types of morpheme: evidence from aphasia, code switching,
and second language acquisition. Linguistics 38: 10531100.
Nelde, P.H. (ed.) (1980). Languages in contact and in conflict. Steiner.
Paradis, M. (ed.) (1987). The assessment of bilingual aphasia. Erlbaum.
Poplack, S. (1980). Sometimes Ill start a sentence in Spanish y termino en espaol: toward a typology
of code-switching. Linguistics 18: 581618.
Poplack, S., D. Sankoff & C. Miller (1988). The social correlates and linguistic processes of lexical
borrowing and assimilation. Linguistics 26: 47104.
Ross, M. (1996). Contact-induced change and the comparative method. In M. Durie & M. Ross
(eds.) The comparative method reviewed: 180217. Oxford University Press.
Salmons, J. (1990). Bilingual discourse marking: code switching, borrowing, and convergence in
some German-American dialects. Linguistics 28: 453480.
Silva-Corvaln, C. (1994). Language contact and change. Clarendon Press.
Stolz, C. & T. Stolz (1996). Funktionswortentlehnung in Mesoamerika. Spanisch-amerindischer
Sprachkontakt (Hispanoindiana II). Sprachtypologie und Universalienforschung 49: 86123.
Thomason, S.G. (1997). A typology of contact languages. In A. Spears & W. Donald (eds.) The structure
and status of pidgins and creoles: 7188. John Benjamins.
(2001). Language contact. An introduction. Edinburgh University Press.
Thomason, S.G. (ed.) (1997). Contact languages: a wider perspective. John Benjamins.
Thomason, S.G. & T. Kaufman (1988). Language contact, creolization and genetic linguistics. University
of California Press.
Volterra, V. & T. Taeschner (1978). The acquisition and development of language by bilingual children.
Journal of Child Language 5: 311326.
Weinreich, U. (1953). Languages in contact. Mouton.

Reconstruction
Derek Nurse

Linguistic reconstruction grew out of nineteenth century work in comparative linguistics, itself an offshoot of an older synchronic tradition. European scholars since
the Roman period had studied Latin grammar. The Renaissance reawakened interest
in Ancient Greek, and awareness grew of Hebrew and Arabic. European expansion
into new areas of the world also brought an increased consciousness of non-European
languages. The greatest, but not the only, single stimulus to historical research
occurred in the late eighteenth century when Western scholars came across Sanskrit,
the classical language of northern India. They quickly recognized the similarities
between it and both the classical and modern languages of Europe. They further
recognized that while such similarity might result from coincidence (or from contact, only later acknowledged), it most likely resulted from genetic relationship,
which implied a common ancestor. This recognition gave rise to the energy and
activity that characterized work on historical and comparative linguistics, work
that continued unabated throughout the nineteenth and well into the twentieth
century. Comparative linguistics here is understood as working backwards in time,
from current or attested languages to an earlier state, whereas historical linguistics
is understood as the study of the development of a language or languages over time,
thus moving forward in time.
While much was achieved, particularly by Germans (e.g. Schlegel, Jakob Grimm,
Bopp) in the first part of the nineteenth century, especially in Indo-European, it was
not until the 1860s and 1870s that the foundation for modern comparative work and
reconstruction was laid by the Neogrammarians (Osthoff, Brugmann, etc.). Their most
striking claim was that sound change proceeeds automatically and without exception:
that is, all instances of a given sound in a given phonetic setting (environment or
context) will change in the same way. Now sounds, hence the shape of words, became
predictable. Once this firm basis for sound change was established, modern reconstruction could start. The fifty years from 1890 to the start of the second World War
saw the production of most of the classic reconstructions of Indo-European or its
subfamilies. Most reconstructions of other language families followed later.
The basic assumptions and procedures involved in reconstruction are easily illustrated via a set of data. Below is a set of words all meaning stool (and by modern
extension, chair) in a set of contemporary Bantu languages. On the left of the line is

216 Derek Nurse

the Standard Swahili form (A), the others (B, C, D, E) are either related languages or
dialect forms of Swahili:

A B C D E
kiti citi kiti shiri kihi

Each of the real words is first segmented, thus A as [k][i][t][i], B as [c][i],[t],[i], etc.
It is then clear that (a) where A has initial [k], B has [c], C [k], D [sh], and E [k]; (b) [i]
corresponds in both positions in all five languages, and (c) A intervocalic [t] corres
ponds to B [t] to C [t] to D [r] to E [h]. By examining many other words in AE we
establish that these three sound correspondences in fact occur regularly and often.
Assuming that AE are related languages/dialects and that all derive from a shared
ancestor language (proto-language, which we will here call Proto-X) to be reconstructed, what might/would the word for stool have looked like? Again proceeding segment by segment, we ask what the initial segment would have been. While in principle
the answer might be [k] or [c] or [sh] or some other consonant, in practice all linguists,
whether familiar with this data or not, would answer *k (see further): any sound or
word preceded by an asterisk is a reconstruction. The initial consonants in languages
AE are said to be reflexes of this *k since they are assumed to derive from it regularly.
So Proto-X is likely to have had a *k and the word for stool started with this *k. Similarly
with [i], except that there is hardly any debate since all five contemporary words have
this [i]. So we assume Proto-X had an *i and it was the first and the second vowel in
the word for stool. Finally, what might/would the second consonant have been? Again,
while in principle it might have been [t], [r], [h], or some other consonant, in practice
most linguists would unerringly agree on *t (see further). So Proto-X probably had a
*t and it was the second consonant in Proto-X stool. Thus the Proto-X word for stool
is reconstructed as *kiti. Such a reconstructed word is known as an etymon or starred
form. The five words above, from languages/dialects AE, are said to be cognate: two
or more words or morphemes are said to be cognate if they are semantically identical
or relatable and can be shown to derive by direct and regular oral transmission from
a single etymon in a proto-language. Kinship terminology is used to describe the relationship among the languages/dialects A, B, C, D, E, and Proto-X: Proto-X is referred
to as the mother, AE are her daughters, and sisters to one another.
In this way, by identifying similar sets of sound correspondences in semantically
identical (or relatable) items across languages, we can build up the probable inventory
of consonants and vowels of a proto-language (similarly, we can also reconstruct nonsegmental features such as tone). From the phonetic level we move to the systemic
level, by ordering the segments and considering them as part of a reconstructed consonant and vowel system. Thence we move to morphology. We in fact already know
that Swahili /kiti/ should be segmented into a stem -ti and a (noun class marker) prefix
ki- and that this is also true for its relatives BE. If we examine more material we find

Reconstruction

many more prefixal ki- (A, C, E), ci- (B), and shi- (D). On this basis we can posit a
*ki- for Proto-X. The more sets we examine the more stems and nominal prefixes we
can reconstruct. Nominal suffixes can also be set up, as can verbal affixes by examining
verbal structures. Once we have reconstructed a set of grammatical morphemes in this
way, we can attempt to put them together in a system and we get a morphology. The
more sets we examine the more we see the outlines of an emerging set of phonological
and morphological alternations. In summary what we can reconstruct is: a probable
consonant and vowel inventory, maybe some of the non-segmental inventory; a lexicon,
whose size is necessarily limited by how much data we have from the daughters; a set
of grammatical morphemes, a morphological system; and something of the (morpho)
phonology. Older books showing reconstructed languages tend to have the components just listed; they are obviously incomplete because of their lack of syntax. More
recent books tend to add a syntactic component but, while it is based on comparison
and does involve reconstruction, it does not follow the steps outlined above. This kind
of reconstruction does not lend itself to syntax.
Certain standard procedures are followed above. One is to choose the reconstructed sound so as to make any changes as natural as possible: the reason most
linguists would automatically choose *k and *t above is because they know of many
other cases where k has become sh or c when in the environment of i (a high front
vowel), and where t has become r or h, but none which work in the opposite direction.
Another is to choose a phonetically natural reconstructed sound and one which offers
as simple a solution as possible. By choosing *k and *t above we only have to account
in both cases for two languages which have changed. Any other choice would involve
more change and more explanation. At least where a proto-language is known or
thought not to lie too far in the past, it is reasonable to think that many of its offspring
will not have changed. A final consideration is what a proto-sound system looks like.
Sounds are not set up in isolation, they form part of a (consonant or vowel) system,
and as each individual member of the emerging system is posited, the comparative
linguist keeps an eye on how they fit in the overall system. All these procedures are
informed by the linguists knowledge of typology universal, genetic, and areal.
Underlying these procedures are certain assumptions. Languages are changing all
the time and at all levels. While most lay people think of change as affecting vocabulary, less obvious but no less pervasive change is also going on in sound systems, morphology, semantics, and syntax. Over a period of time a language will change so much
that a new language can be said to have arisen. This can be seen in any language for
which lengthy written records are available: Sanskrit and its descendants in northern
India; Chinese; several languages in western Europe. This, taken together with the fact
that language communities split and separate, is the basis of the genetic model and
thus for the comparative method and reconstruction. At the same time we assume that
over a long period of time, if one instance of a sound in a given environment changes,

217

218 Derek Nurse

then nearly all the other instances of that sound will change in the same way. Whereas
the Neogrammarians predicted that all instances would change, we have learned in the
meanwhile to accept a modified version of this hypothesis because counter-changes
obliterate or modify the effect of a particular change in some words, and because language does not exist independently of social forces, which lead to the adoption of one
competing pronunciation over another, to the puzzlement of a comparativist looking
at the phenomenon some centuries later. We also assume that some kinds of sound
change are more likely, and thus to be preferred as explanatory models, than others,
because we have the accumulated wisdom of more than a century to rely on.
What is described in the previous three paragraphs is the operation of, and the
procedures and assumptions behind what is called the comparative method, which is
usually defined as the working out of regular phonological correspondences between
two or more languages with the aims of (1) establishing that the languages are related,
and (2) then reconstructing features of their ancestor. What has been done above is
(2), not (1), as, for purposes of demonstration, it is taken for granted that AE are
related. If instead we had added to the set other words, such as English stool, German
Stuhl, Dutch stoel, we get two impressions: first, these words might belong together, and,
second, they do not belong with AE because they are phonologically incompatible
with them. Both these impressions can be corroborated by examining more sets from
English, German, and Dutch. We could then treat the latter as we have treated AE
and reconstruct features of the language ancestral to English, German, and Dutch
(Proto-Germanic). The proto-languages ancestral to AE and to English-GermanDutch are not the same.
Proto-languages are usefully seen as being of two types: intermediate and ultimate.
Thus, if we examine all the several hundred Bantu languages and apply the methodology
above, they will fall into many small subsets (so AE, FH, IP, etc.), on the basis of
shared features of vocabulary, sound, or morphology. The members of each subset
can be derived from a protolanguage, which is an intermediate stage, because if all
these proto-languages are in turn juxtaposed, the same procedure can be applied to
them, and ultimately they will be assignable to a single proto-language (Proto-Bantu).
Similarly, the intermediate proto-languages, Proto-Germanic, -Romance, -Slavic, and
others can all be grouped under a single ancestor, called Proto-Indo-European. This is
the family tree model of genetic relationship. Languages assigned to different ultimate
proto-languages are members of different language families. A very few proto-languages
(e.g. Proto-Romance) can be checked against comparable historical languages actually
attested (post classical colloquial Latin), but most cannot.
There is a wide range of opinions among linguists as to the reality of reconstructions.
All linguists share the minimal view that starred sounds such as *t or *k above are at
least formulae to account for sets of correspondences between related languages. Some
linguists would not want to go much further than this and would deny phonetic reality

Reconstruction

to these reconstructions. At the other end of the scale are linguists so confident of the
reality of their reconstructions that they write stories in them or craft cultural history
based on reconstructed vocabulary. A sensible middle view is that, while allowing for the
possibility that new data or improved methods might alter our reconstructions, someone who reconstructs is advancing a hypothesis about history and that our reconstructions are likely to be largely accurate, provided we work systematically. This can be
demonstrated by reconstructing Proto-Romance from contemporary Romance languages and then comparing it with Latin: the two are largely identical. Reconstruction
is a matter of probability. We construct a hypothesis on the basis of current data and
methods and accept it until something happens to make us change our mind the
typical and normal use of theoretical method.
The degree of probability rests on certain variables. A major variable is time. Intermediate proto-languages, being projections from contemporary or attested languages,
are located in the relatively recent past and we can be relatively confident about their
reality. Ultimate proto-languages represent a stage much further in the past. We must
have more reservations about their reality because many of the languages on which
they would ideally be based have died out, and because the passage of time has so
corroded phonological patterns that they are hard or impossible to discern.
The question of the degree of probability of reconstructions lies at the center of a
current controversy. All comparativists would agree that reconstruction is possible of
proto-languages up to 5,000 (e.g. late Proto-Indo-European), 7,000 (early Proto-IndoEuropean) or even 10,000 years ago. Beyond that they do not agree. Archaeologists
posit that humans have existed for several million years and that humans capable of
speech have existed for tens of thousands, maybe hundreds of thousands, of years.
A group of (mainly) Russian scholars has been active for decades trying to combine currently accepted language families and reconstruct larger language groupings
that would project backwards far enough in time to fill in the gap between currently
accepted groupings and the likely beginning of human speech. This activity has spread
in the 1970s and 1980s, mainly to some American scholars. Other scholars are content
to sit and watch this activity in scepticism because they feel it mixes the comparative
method with too large a dose of unsubstantiated resemblance.
Three other possible roads lead to the future. One lies in extending reconstruction.
Many intermediate proto-languages in commonly accepted language families have not
yet been adequately reconstructed (e.g. Proto-Germanic). Similarly, the majority of the
worlds languages have been arranged into tacitly accepted families but reconstruction
of their proto-languages has not been consistently carried out. Another lies in broadening what is reconstructed. Despite what is said above, most existing reconstructions
have been of sound inventories and lexicon. There is much current synchronic interest
in verb systems, especially in tense, aspect, and associated categories. Reconstruction
of such morphosemantic categories is as yet largely unexploited. Thirdly, scholars are

219

220 Derek Nurse

just starting to explore the use of computers in reconstructions. Computers can and
are (e.g. in Siouan and Proto-Baltic) being used for keeping record of what is being
reconstructed by conventional methods. But they can also be programmed to actually
search large bodies of data for sound correspondences, the basis for reconstruction.
So far this has only been used for Proto-Algonquian but is currently being applied to
parts of Sino-Tibetan and a new reconstruction of Bantu. Against these exciting possibilities must be set the fact that the small number of linguists trained and presently
working in comparative methods does not allow for large scale activities.

References
Hoeningswald, H.M. (1960). Language change and linguistic reconstruction. University of Chicago
Press.
Labov, W. (1981). Resolving the Neogrammarian controversy. Language 57: 267308.
Meillet, A. (1967). The comparative method in historical linguistics. Champion.
Osthoff, H. & K. Brugmann (1878). Preface to: Morphologische Untersuchungen auf dem Gebiet der
Indogermanischen Sprachen 1. In W.P. Lehmann (ed.) A reader in 19th century historical IndoEuropean linguistics. Indiana University Press.

Register
Norbert Dittmar

1. History of the term register

Firths (1968 [1957]) speech-theoretical perspective is at the root of the term register in
modern linguistics. In contrast to the formalistic line of American structuralism, Firth
postulated that linguistic signs relate directly to the context, i.e. their usage depends on
the context-of-situation. Firths examples are often from the realm of restricted languages: the language for regulating air traffic, weather forecasts, congratulation cards,
and in simplified registers of game instructions (chess, bridge). In these examples, the
stable connection between context and linguistic-communicative patterns is evident.
Firths aim is an integrated theory of meaning in which the speech users and their
speech usage are given equal status. The term register was introduced by the Firthian
linguist Reid (1956). The first conceptual and theoretical elaboration of the term can
be found in Halliday, McIntosh & Strevens (1964), a book on applied linguistics. The
basic idea underlying this essay, which is to base the notion dialect on the dimension
speech user and the notion register on the dimension speech usage, is described fully in
Hallidays Language as social semiotic (1978). The term register is one of the fundamental pillars of the Hallidayan conception of sociolinguistics.
Halliday distinguishes between dialect as a variant definable in terms of speech
users and register as a variety of speech usage. The dialect is what a person speaks,
determined by who he is; the register is what a person is speaking, determined by
what he is doing at the time (Halliday 1978:110). There are many attempts to define
register in the course of the twelve chapters of Hallydays book. A typical statement
is the following:
A register can be defined as the configuration of semantic resources that a member of
a culture typically associates with the situation type. It is the meaning potential that is
accessible in a given social context. Both the situation and the register associated with
it can be described to varying degrees of specificity; but the existence of registers is a
fact of everyday experience speakers have no difficulty in recognizing the semantic
options and combinations that are at risk under particular environmental conditions.
Since these options are realized in the form of grammar and vocabulary, the register is
recognizable as a particular selection of words and structures. (1978:111).1

. The linguistic consequences of this concept bear on the definition of varieties, which is discussed in Section2. The term register has to be distinguished from code, dialect, and style.

222 Norbert Dittmar

According to Halliday, linguistic-communicative registers are lexico-grammatical

varieties related to different contexts of situation. This is why types of speech situations
have to be distinguished; they vary in three respects (Halliday 1978:31):
First, what is actually taking place; secondly, who is taking part; and thirdly, what part
the language is playing. These three variables, taken together, determine the range
within which meanings are selected and the forms which are used for their expression.
In other words, they determine the register.

(Speech-)situation types are determined by linguistic-communicative patterns of the

discourse field, by (discursive) tenors, and by the discourse mode. These three components activate certain networks of semantic options (1978:123), i.e. semantic choices,
which lead to the linguistic linking of an action pattern. Halliday (1978: 123) calls
this text-specific process of semantic selection the range of meaning potential. Three
parameters determine the linguistic forms of a register:
i. Field (discursive speech-usage field). With different types of action our language varies. Different lexico-grammatical patterns go hand in hand with differences in action
patterns. Part of our actions is its subject matter; thus, the discursive speech-usage
field includes the topic as well. Therefore, the discursive field is first and foremost an
action field. During the action football, with football as the topic of conversation, it is
possible to speak about the weather without having to call the current action meteorology. Nevertheless, the verbal action during the game has to be distinguished from
the discussion of football in the pub. The difference is expressed via the concept discourse field.
ii. Tenor (discourse style). Tenor refers to the degree of formality between the
interaction partners. In other words, role relationships are associated with different
style characteristics on a scale of formality. Examples include typical institutional
role relationships such as that of teacher and pupil, parents and children, children
in their peer groups, doctor and patient, salesperson and customer, people meeting
in a train, etc. The generic label is role relationship, whereas institutional roles are
special forms which are as a rule clearly distinctive. Halliday qualifies institutional
role relationships (including also categories such as lectures, church services, and
cocktail parties) as the stabilized pattern of the tenor of discourse (1978: 222).
Inpragmatic research, this would be described as communicative usage patterns to
which a pattern knowledge can be assigned.
iii. Discourse mode. This term refers to the linguistic shaping of the discourse, e.g. the
written medium is to be distinguished from the oral form. Another central issue is
what function the language is supposed to fulfill in the social context. A particular
rhetorical channel (1978: 222) is chosen for teaching or poetry, in advertising,

Register

to commission something, or to give orders. Functions of the narrative have to be

distinguished from instructions, etc. In this context, questions also arise as to
whether language is used to convince, to sell, to control, to explain, etc. In addition, it
would be part of a description of the discourse mode to see whether utterances are
organized in an ego-related or an alter-related way, i.e. with emphasis on the speaker
or on the hearer.
The term register is borrowed from organ playing. Language is seen to be regulated in
the same way as the musical tuning of an organ: the linguistic-communicative behavior
is thus situationally and context-specifically calibrated. To sum up, register in Hallidays
sense means congruence between situational features, a discourse mode (the medium
of the interaction), a level of intimacy and emotional state (tenor), and social roles that
are involved in the communication process (institutional and group-specific role relationships). On a linguistic level the congruence between these parameters is expressed
by various linguistic means. These essential features of Hallidays conception, in spite
of criticism on specific points, have been handed down to this day as the core of the
sociolinguistic term register. The term register has not, however, found its way into
the wide domain of pragmatic research on institutional communication.
The term register is widely known in the anglophone linguistic world. A variation on the denotation of the term, focusing on the receiver, was developed in the
United States by Charles Ferguson. Role constellations, the corresponding functional
appropriateness of the communicative code (discourse mode), and the situation are
the parameters used by Ferguson to distinguish between a baby register (motherese),
a foreigner register (foreigner talk), and the register of the sports announcer (sport
announcer talk) (Ferguson 1977). Consider the typical asymmetrical situation of
parents and their children, or of foreigners and native speakers, as the parents (with
the fully developed linguistic competence of adults) communicate with the languagelearning child (say, three to four years old), or a native speaker communicates with a
non-native speaker. Because of the differences concerning the lexical range, knowledge
of grammatical rules, etc., a form of accommodation takes place: the linguistic and
grammatical means are radically simplified, and other areas of the linguistic code are
over-articulated to facilitate understanding. Clearly, the registers have a situational
and role-specific relevance. Fergusons empirical and theoretical work on his original
understanding of the term register found approval in the linguistic and sociolinguistic
literature. In a recent extensive empirical study, however, Jrg Roche (1987) rejects
Fergusonss interpretation of register as too wide, too vague and too undifferentiated
for the domain of the foreigner register. Roche finds a great deal of variation in the discursive forms of foreigner registers; this induces him to put the situational and basic
parameters at the centre of his work.

223

224 Norbert Dittmar

Finally, the francophone usage of the term register should be mentioned. Ager
(1990) and Sanders (1993) offer an overview. The French term registre was coined in
the 1970s and distinguishes so-called niveaux de langue: populaire, familier, courant,
soutenu, acadmique/littraire. These levels in Hallidays terminology tenor/discourse
style can be differentiated according to the degree of formality of the speech situation. At the same time, the degree of formality has to do with social and communicative roles in the interaction, with a non-standard pronunciation, and the use of
non-standard rules. Grammatical deviations from standard French make up vulgar
French (the lowest level). The niveau familier, suggesting familiarity between interlocutors, is the everyday language between intimate friends (e.g. one would say Tu
peux me donner un conseil? instead of Peux-tu me donner un conseil?). The niveau
courant represents everyday or colloquial usage. The question form in this style would
be Est-ce-que tu peux me donner un conseil? In the sophisticated, more pretentious
speech style (soutenu), inversion is necessary: Peux-tu me donner un conseil? The
highest level, academic and literary language usage, demands certain conventional
formulae derived from classical French. The most striking feature of the French notion
of register is its definition along the formal-informal continuum, related to grammatical and especially lexical differences. Sanders (1993:33) offers a summary of different features. In French sociolinguistic research, the interference of the concept of
register with sociological concepts such as age, social class, group, and institutional
vs. everyday situations is rarely clarified. The more recent publications by BlancheBenveniste (1990) and Deulofeu (1986) have shown that detailed linguistic descriptions of the linguistic and grammatical means of the French colloquial language are
a prerequisite for a meaningful explanation of the register concept. Thus, the French
register research puts the discourse style (tenor) at the center of description at the
stylistic level, whereas the sociological correlates are not explored.

2. Systematization of the term

According to Sanders (1993:2753), there is a correlation between non-linguistic
parameters such as social characteristics of the interactants, the context of the situation,
the discourse modes or communication contexts and the linguistic/communicative
features of the spoken language, which are defined by these parameters. The notion of
register captures this correlation. Figure1 shows the connection between these factors:
the age, sex, social class, etc. of the speakers and hearers in interaction with the context
of the situation and the discourse modes (context of communication) cause linguistic
register forms or socio-situational variations.

Register

LOCATION
age
sex
social class
education
geog. provenance/social status
regional affiliation
Political
social
persuasion
attitude

CONTEXT OF
SITUATION
social roles
setting and location
subject-matter

RELATIVE
AGE
SEX
SOCIAL
CLASS,
ETC.

INTERLOCUTOR 1 n
age
sex
social class
education
geog, provenance/social status
regional affiliation
Political
social
persuasion
attitude

CONTEXT OF
COMMUNICATION

R
E
G
I
S
T
E
R

medium (speech/writing/
electronic)
conditions (space; time;
preparedness)
social function of speech event
linguistic context

Figure 1. Interaction between the register-variation defining parameters

The congruence of the linguistic levels of a register results in the interaction of at

least four factor fields (see Figure1). The varieties dialect and sociolect may obviously
overlap. Therefore, one of the urgent problems in modern research is the differentiation of the terms variety and register.
The examination of spoken language by Gadet (1989), Blanche-Benveniste (1990),
and Deulofeu (1986) supplies differentiated descriptions of the phonetics/phonology
and morphosyntax of the French language in everyday situations of language usage.
Only with the help of such extensive variety descriptions can registers be linked productively with non-linguistic factors.
According to Arriv, Gadet & Galmiche (1986:597600), registers are determined
on the level of phonetics (e.g. liaisons, harmonization of vowels), prosody (intonation,
pauses, and slowing down), lexicon (synonyms), and syntax (relative pronouns and
interrogative clauses). Looked at in particular from the point of view of systematic
characteristics of the spoken language, the division of registers into niveaux de langue
seems questionable to these authors: langue populaire has as its criterion social class;
langue familire refers to situation.
While registre is an important sociolinguistic term in French research, in German
research the term communicative style is generally used for similar phenomena. Following the ethnographic work of John Gumperz and using the notion of contextualization

225

226 Norbert Dittmar

cues, styles are seen as congruences between group-specific factors, emotional states of
being (discourse modes), conditions of situational frame, and linguistic means (lexical
and grammatical choices, prosody). Examples of this kind of work on communicative
styles are Hinnenkamp & Selting (1989) and Dittmar (1995a, b). Here, in a cautious
approach towards a possibly more formal definition, style is understood in terms of
a number of basic pragmatic parameters (situation, relationships between interacting
people, aims and purposes of action, institutional and social roles, etc.). In contrast to
the francophone and the anglophone notions of register, the leading idea in German
research is that styles change dynamically in the course of interaction, or are subject
to change. The pragmatic factors which affect different levels of style are interpreted
as dynamic, changing constellations in the course of their interaction. Germanophone
sociolinguistic research takes up the more radical problem of isolating the pragmatic
parameters that control the dynamics of the communication in the course of the interaction and that considerably influence, thereby, the sequentiality in a discourse, the
choice of linguistic means, and the synchronization of interactants in conversations.
When we ask what value the term register has for sociolinguistic or pragmatic
research, it is important to embed the term into a sociolinguistic theory or into variation linguistics. Since such an embedding is suggested in Halliday (1978), I will bring
it up for discussion in the following. There are two questions:
i. What relationship does register have with other varieties in the variation
continuum?
ii. What relationship does it have with the social structure?
According to Gregory (1967), dialectal varieties are marked by social features of the
speakers/users in situations of usage. In contrast, diatypical varieties are marked by
features of the usage of speakers in situations. The difference between diatypical and
dialectal varieties lies in the relationship with the situation, the context, and the style
levels (see the table in Halliday 1978:225).
Halliday (1978: 35) adopts Gregorys (1967) differentiations. Figure 2 contains
the essential points for distinguishing between the variety dialect and the variety
register. One point is that dialect is associated with habits (habitus), whereas register
involves social processes (e.g. the social division of labor) referring to the situation.
Note that the variables that control dialect are, for instance, social class, rural or
urban background, age and generation, or sex; in contrast, the control variables of
register are discourse field, discourse style, and discourse mode. A further socialpsychological point is that dialects differ with regard to the evaluations which are
associated with them; registers are characterized by the differences between spoken
and written, and between language embedded in actions vs. language in the service
of reflection.

Register
Dialect (dialectal variety)
= variety according to the user

Register (diatypic variety)

= variety according to the use

A dialect is:
what you speak (habitually)
determined by who you are (socio-region
of origin and/or adoption), and
expressing diversity of social structure
(patterns of social hierarchy)

A register is:
what you are speaking (at the time)
determined by what you are doing (nature
of social activity being engaged in, and
expressing diversity of social process
(social division of labour)

So in principle dialects are:

different ways of saying the same thing
and tend to differ in:
phonetics, phonology, lexicogrammar (but
not in semantics)

So in principle registers are:

ways of saying different things
and tend to differ in:
semantics (and hence in lexicogrammar,
and sometimes phonology, as realization of this)

Extreme cases:
antilanguages, mother-in-law languages

Extreme cases:
restricted languages, languages for special purposes

Typical instances:
subcultural varieties (standard nonstandard)

Typical instances:
occupational varieties (technical, semi-technical

Principal controlling variables:

social class, caste; provenance (rural/urban):
generation; age; sex

Principal controlling variables:

field (type of social action); tenor (role
relationships); mode (symbolic organization)

Characterized by:
strongly-held attitudes towards dialects as
symbol of social diversity

Charaterized by:
major distinctions of spoken/written; language
in action/language in reflection

Figure 2. Varieties: The difference between dialect and register

social system
adult linguistic system:

family
role
systems

status/role
relationships
codes
(principles
of semiotic
organization)

situation types or social context,

incl. critical socializing contexts
field tenor mode

social dialect

register

text in situation

Figure 3. Schematical representation of the interaction between social structure

and language structure

textual

context of culture

semantics:
functional
components:
interpersonal

hierarchy,
caste,
social class

lexicogrammar,
phonology

ideational

social structure

227

228 Norbert Dittmar

Figure3 (adopted from Halliday 1978:69) illustrates the connection that exists
between social systems and texts in situations. This leads to a further differentiation
between codes (principles of semiotic organisation of symbols, which are determined
by social class, family role systems, and status role relations), social dialects (determined both geographically and by social class), and register (governed by cultural and
social context and situation types).
Codes are not varieties of a language on a par with registers or dialects, but abstract
semiotic systems (see Halliday 1978:68). Codes are behavior systems which contain
verbal and non-verbal modes of behavior and they are made topical in the language
usage by registers. This is why codes as organization principles of sign systems are
superordinate to registers used in actual situations.
3. Perspectives
Nabrings (1981) discusses the register concept from the point of view of current
methodological approaches for the recording of intra-linguistic variability.2 The primary categorization dimension for register, then, is situational variation (diaphasic),
although it has to be seen as expanded by diastratic factors (social roles), as in Halliday
(1978). As a variety according to use, it is in accordance with the intra-linguistic criteria
of its definition, closely related to functional styles. While linguists who examine the
functional styles say that there is a causal connection between the function of language and the speech form developed for this (Nabrings 1981:200), the connection
postulated by British researchers between linguistic registers and application situations
is conventionally established. The functional-stylistic components of the usefulness of
a variety, or of specific linguistic means for a discourse topic, are rather regarded as
secondary definition features. The key to understanding this historical, school-specific
difference lies in the Firthian tradition of functional linking of linguistic utterances
with the social context. As Halliday, McIntosh & Strevenss (1964) discussion shows,
Firths approach was closely linked with applied linguistics; they have developed the
term register because as a term it lent itself to the description of their observations
that foreigners who learn a second or foreign language in certain interactions with
native speakers make context-related mistakes or use styles that are inadequate. From
the errors manifested in written and spoken discourses one can infer that registers
are the expression of a close connection between speaker, situation, and language use.

. Nabrings, following Coseriu, isolates four dimensions of linguistic variation: the diachronic, the
diastratic, the diatopical, and the diaphasic; with the help of these four dimensions, she develops a
coordinate system which makes it possible to assign specific varieties to these dimensions.

Register 229

Such investigations by Halliday and others have resulted in the so-called communicative
approach to foreign language teaching. Applied linguistics has thus provided data to
substantiate the notion of register. The basis for this kind of research is the understanding of registers as conventional usage patterns defined by roles, situations, and
discourse topic. The three branches of research presented in the following explicitly or
implicitly deal with the description of registers, but they differ in that they define the
dimensions speaker qua social roles, situations qua institutional or group-specific constellations, and theme/discourse topic qua field of action either as connected, indivisible
components of interaction, or as isolable factor clusters. In the domain of applied linguistics or language teaching, the experience-oriented relevance of works on register
has been accepted without problems. Theory-relevant reflections have been made in
the following four domains: functional linguistics la Ferguson, intercultural communication, linguistic pragmatics, and variation linguistics.
3.1 Functional linguistics la Ferguson
The interplay between interaction roles, situation, and discourse topic as an explication of communicative needs characterizes the functional approach by Ferguson. The
foreigner register (foreigner talk) is the specific interactive style between native and
non-native speakers, which in a similar form exists between parents and their children
(baby register). Ferguson isolated the principle of the interaction of the three parameters
of role, situation and topic, using as his example the sports announcer register.3
The features of the foreigner register, for example, can be formulated following
taxonomic principles (see Heidelberger Forschungsgruppe 1975). In certain situations, an understanding between native and non-native speakers can only be achieved
through more or less extreme accommodation in the form of a special register. Even if,
up to now, the establishment of universal characteristics for registers has not been possible, there are typical recurring procedures. Especially in applied linguistics (foreign
language teaching), a continuum of a strong-to-weak shaping of the foreigner register
was recorded (see Rost 1989). Fergusons essays have triggered a number of investigations in the domain of childrens acquisition of language, as well as in that of second
language acquisition or foreign language teaching.
. The speech situation is the following. The reporter sees the match before him/her and tries to
give a verbal account of it in such a way that the listener can comprehend the main moves of the
match without actually watching the match. In order to give a linguistic account of the match, the
reporter uses the informal medium of the spoken language. At the same time, an important feature
of his/her discourse is the time pressure exerted by the sports event itself. This role constellation
calls for linguistic adaptations: the reporter has to verbally present what s/he is observing in such
a way as to enable the listener to paint a lively picture of the sports event. In the verbal utterances,
the complexity of the activities should be portrayed as accurately as possible.

230 Norbert Dittmar

In his standard work on structure and variation in the German used by Germans
when dealing with foreigners, Roche (1989) chooses the term xenolect instead of foreigner register. Roche takes dynamic-pragmatic concepts of description as a basis for
his analysis. He discovers linguistic grading in the xenolects and connects this with
the kind of communicative tasks (the purpose of the discourse) and the social roles
of speaker and hearer. By examining the different factors, isolated according to their
functions, for the dimensions role, discourse topics, and social contexts, he does
not follow Hallidays or in a wider sense Fergusons (typical structuralist) approach,
although his starting point is indeed Fergusons work.
3.2 Intercultural communication
Using ethnographic and conversation-analytical methods, John Gumperz (e.g. 1982)
has concerned himself, within the framework of his theory of contextualization, with
failures in intercultural communication. Gumperzs observations do not relate to the
contexts of foreign language teaching or second language acquisition. He concentrates
on the communication between native speakers and non-native speakers although
usually the non-native speakers have been living in the host country for quite a long
time and are, as a rule, very advanced speakers of the target language. Because of contrasts between values and norms in the mother tongue and the target language, different
pragmatic pattern knowledge, different routines in the usage of gestures and prosody,
different communicative formulae and speech usage patterns may lead to misunderstandings. From the uncovering of misunderstandings and failures in the communication (through qualitative ethnographic descriptions), one can infer specific speech
usage patterns or register characteristics (e.g. during job interviews). Especially during
job interviews and councelling sessions, essential characteristics have been uncovered
through a number of misunderstandings, which allow for a relatively detailed characterization of the underlying norms and their linguistic facets as registers; nevertheless,
the term register is not used in this research.
3.3 Linguistic pragmatics
In a number of works Rehbein (partly in cooperation with Ehlich) examined communicative patterns in institutional communication. (See, e.g. Rehbein 1983.) The
interactions between teachers and pupils; judges, the defence, and witnesses during
a trial; foreigners and locals in work processes; and physicians and patients in the
hospital, have been described. To be able to do this, the researchers relied upon the
social/institutional roles and the concrete situation in the process of communication,
investigating certain forms of sequences and speech usage patterns. The difference

Register

between these works and those that use the term register is that they do not operate
with taxonomic features, but describe the dynamics of the on-going communication
situations pragmatically (via the interaction of different contextual, situational, and
personality-related parameters). A central term in these investigations is pattern
knowledge (Musterwissen), which represents the basis for the realization of speech
in the discourse or interaction. The term pattern knowledge refers to cognitive processes (the psycholinguistic basis), whereas register descriptions are seen as surface
descriptions. Pattern knowledge and register can converge in meaning, though, since
the register always involves the dimension of pragmatic knowledge. Typical institutional interaction patterns are isolated by Rehbein (and others working in the domain
of institutional communication). The behavior is not structuralistically described by
given parameters, but an open system accessible only to observation. Relevant literature on this domain is to be found in Dittmar (1995c).
3.4 Variation linguistics
Variation linguistics (see, e.g. Hudson 1980; Sanders 1993; in German Variettenlinguistik, see Nabrings 1981) stresses the need to explain how registers are defined
by the interaction of the diaphasic and the diastratic dimension. Blasco Ferrer (1990)
sees degree of formality as the decisive characteristic feature of the register concept.
For him, the continuum italiano populare italiano standard belongs to a grading of
registers. With the help of registers, he tries to isolate group languages (not to be confused with sociolects) on a wider scale. Nevertheless, his definitions would be totally
unacceptable for Halliday, who distinguished between the habitus-usage of varieties
(e.g.dialects) as outer markings of the geographical or social identity, and the necessary, situation-dependent, institutional and other speech usage patterns, which above
all show drastic semantic differences. Ferrer thus seems to come closer to Labovs
(1966) notion of style than to Hallidays basic idea of register.
Other variation linguists are interested in the term register and make reasonable
use of the term and the related instruments of description, in particular for a rigid
description of different forms of the spoken language. Blanche-Benveniste, Deulofeu,
Schlobinski, Selting, and others examine syntactic and semantic patterns of the spoken
language that go along with the usage of conjunctions, modalizations, definite verbs,
thematization, etc.; they only tentatively include the various discourse types and social
contexts. The predominant argumentation is that a detailed description of the regular
variation of the spoken language is necessary in order to establish, in the next phase of
analysis, the connection with situations and roles in the interaction.
A question occupying Berruto (1995) is what place registers could have in the
structure of varieties isolated above. In particular, how can we resolve the problem that

231

232 Norbert Dittmar

registers simultaneously belong to diaphasic and diastratic varieties? He discusses two

possible, and interesting, perspectives; both suggest that the dimensions are hierarchially
connected with each other and do not have the same status:
a. Studying the acquisition of registers would give interesting information. The
hypothesis is that the child first acquires the diatopical dimension, then the diastratic,
and only then can s/he learn the situation-specific differences (register). If this is the
case, then it seems reasonable to formulate a hierarchic relation between diatopical
> diastratic > diaphasic.
b. A second possibility was formulated by Wunderli (1992). Wunderli views variation
as modularly organized. There is a module of the primary manifestation level which
refers to the user of language, consisting of two primary axes, the horizontal diatopical and the vertical diastratic. The primary manifestation level is differentiated along
the parameter of the communicative situation (the diaphasic dimension). Wunderli
structures his second module on Hallidays three parameters: discourse topic (field),
discourse mode (understood in a slightly wider sense than in Halliday), and discourse
style (Hallidays tenor; in modern sociolinguistic research the difference between
formal and informal styles). According to Berruto, these, altogether five, dimensions
allow a classification of speech productions on the two levels of the modules.
The possibility of an integration of the given levels (a hierarchic structure) is definitely
worth looking into. What is not solved in Wunderlis suggestion is the internal structure of the module diaphasic dimension. It is indeed urgently necessary to carry out
pragmatic, fundamental research that investigates the following interactions in more
detail: speech usage and situation types; speech usage and social/institutional roles;
speech usage and different moods/emotions of the speakers.

References
Ager, D. (1990). Sociolinguistics and contemporary French. Cambridge University Press.
Arriv, M., F. Gadet & M. Galmiche (1986). La grammaire daujourdhui. Flammarion.
Berruto, G. (1995). Fondamenti di sociolinguistica. Editori La Terza.
Blanche-Benveniste, C. et al. (1990). Le Franais parl. CNRS/Didier Erudition.
Blasco Ferrer, E. (1990). Italiano populare a confronto con altri registri informali. In G. Holtus &
E.Radtke (eds.): Sprachlicher Substandard III: 211243. Narr.
Deulofeu, J. (1986). Syntaxe de que en franais parl et le problme de la subordination. RFP 8.
Dittmar, N. (1995a). Theories of sociolinguistic variation in German. In P. Stevenson (ed.) The German
language and the real world: 135168. Oxford University Press.
(1995b). Sociolinguistic style revisited. In J. Werlen (ed.) Verbale Kommunikation in der Stadt:
111134. Narr.
(1995c). Studienbibliographie Soziolinguistik. Groos.

Register

Ferguson, C. (1977). Simplified registers, broken language and Gastarbeiterdeutsch. In C. Molony,

H.Zobel & W. Stlting (eds.) Deutsch im Kontakt mit anderen Sprachen: 2539. Kronberg.
Firth, J.R. (1968). A new approach to grammar [1957]. In Selected Papers of J.R. Firth, 19521959:
114125. Longman.
Gadet, F. (1989). Le Franais ordinaire. Colin.
Glser, R. (1976). Die Stilkategorie Register in soziolinguistischer Sicht. Zeitschrift fr Phonetik,
Sprachwissenschaft und Kommunikationsforschung 29: 234243.
Gregory, M. (1967). Aspects of varieties differentiation. Journal of Linguistics 3: 177198.
Gregory, M. & S. Caroll (1978). Language and situation. Routledge & Kegan Paul.
Gumperz, J. (1982). Discourse strategies. Cambridge University Press.
Halliday, M.A.K., A. McIntosh & P. Strevens (1964). The linguistic sciences and language teaching.
Longman.
Halliday, M.A.K. (1978). Language as social semiotic. Edward Arnold.
Hartung, W. & H. Schnfeld (eds.) (1981). Kommunikation und Sprachvariation. Akad. der Wissenschaften
der DDR, Akademieverlag.
Heidelberger Forschungsgruppe (1975). Sprache und Kommunikation auslndischer Arbeiter. Scriptor.
Hinnenkamp, V. & Selting (eds.) (1989). Stil und Stilisierung. Niemeyer.
Hudson, R.A. (1980). Sociolinguistics. Cambridge University Press.
Labov, W. (1966). The social stratification of English in New York City. Center for Applied Linguistics.
Nabrings, K. (1981). Sprachliche Varietten. Narr.
Rehbein, J. (1983). Komplexes sprachliches Handeln. Narr.
Reid, T.B.W. (1956). Linguistics, structuralism, philology. Archivum Linguisticum 8.
Roche, J. (1989). Xenolekte. de Gruyter.
Rost, M. (1989). Sprechstrategien in freien Konversationen Narr.
Sanders, C. (ed.) (1993). French today. Cambridge University Press.
Spillner, B. (1987). Style and register. In U. Ammon, N. Dittmar & K. Mattheier (eds.) Sociolinguistics,
Vol 1: 273285. de Gruyter.
Wunderli, P. (1992). Le problme des varits diastratiques. Communication and Cognition 25: 171189.

233

Typology
Bernard Comrie

The term typology is used in a narrower and a broader sense in linguistics. In the
broad sense, which will be adopted here, the typological approach to linguistic investigation is characterized by the following features:
1. it draws on data from a wide range of languages;
2. it is data-driven rather than theory-driven.
In common with many other approaches to linguistic investigation, including for
instance generative grammar, the typological approach aims to discover general
properties of human language, the constraints on variation across languages, and the
principles that govern permitted variation across languages. The typological approach
is sometimes called language universals and typology, giving due emphasis both to
the search for general properties of language and to the principled study of crosslinguistic variation. Some linguists prefer to define both language universals and language typology in narrower senses, restricting language universals to the search for
those properties that are common to all languages, and restricting language typology
to the search for the basic parameters that govern cross-linguistic variation. However,
whatever the terminology adopted, the overall goal remains the same: the recognition that there are both significant similarities and significant differences among
languages, that these differences can, at least to a considerable extent, be given a
principled account, and that such a principled account must be empirically based
on reliable data from a range of languages. General works surveying this approach
include Comrie (1989) and Croft (2002).
Linguistic typology, though having antecedents in, for instance, the work of
Wilhelm von Humboldt in the early nineteenth century (concentrating mainly on
morphological typology) and Prague School linguists in the earlier twentieth century,
can date its current upswing to the work of Joseph H. Greenberg on word order
typology. (Other significant contemporary schools are to be found in Germany, in
particular Hansjakob Seiler and his former students, e.g. Christian Lehmann, and
in Russia, especially in the St. Petersburg (Leningrad) typological school founded by
Aleksandr A. Xolodovi and continued by, among others, Vladimir P. Nedjalkov).
Greenberg (1966) is a seminal article in that it demonstrated how progress could
be made using a broad sample of languages in order to investigate cross-linguistic
restrictions in a particular linguistic domain, in this case word order (more accurately:

Typology

constituent order). This article by Greenberg defined a number of concepts that have
come to be accepted in typological research, even if not always with the same terminology. An important distinction is made between implicational and non-implicational
universals. An implicational universal establishes a relation of the type if p then q
between two linguistic properties, for instance if a language has first-person reflexive
pronouns, then it has third-person reflexive pronouns. An implicational universal in
fact combines, in the narrower senses, universals and typology, since it (1) excludes
one of the four combinations of p and q, namely p & not-q (in the example given,
a language having first-person reflexive pronouns but lacking third-person reflexive
pronouns), thus formulating a universal restriction, while (2) providing a classification of languages into the three permitted types (p & q, not-p & q, not-p & not-q).
A non-implicational universal is one that does not relate properties in this way, for
instance all languages have some degree of hierarchical syntactic structure. A second
important distinction is between absolute universals, i.e. those that hold without
exception for all possible human languages, and universal tendencies (also called:
statistical universals, distributional universals), i.e. those that hold for the vast majority of human languages, to a greater degree than would be expected by chance, but
which nonetheless do have exceptions. Examples of universal tendencies include:
(1)In the basic order of elements within the clause, it is much more usual for the
subject (S) to precede the object (O) than vice versa (indeed, probably less than 1%
of the worlds languages have OS rather than SO as their basic word order);
(2) If a language has basic VSO word order (where V = verb), then it will have
prepositions rather than postpositions (a generalization which has a small number
of exceptions). Note that of these examples of universal tendencies, the first is nonimplicational, the second implicational.
The question of the data-base for typological research is an important one. The
aim of this research is to characterize properties of the human language potential.
However, the data available are necessarily taken from languages that are accessible,
i.e. at least in principle, all languages currently spoken and all extinct languages for
which we have sufficient data although in practice the lack of documentation of
many languages substantially reduces this availability, and the ongoing phenomenon
of accelerated language extinction threatens to eliminate the possibility of a sufficiently
broad data-base. The most important point is to guard against biases that would lead to
over-emphasis on languages that are related to one another. An obvious bias is genetic
relationship, i.e. selecting a number of languages that are genetically closely related to
one another has the danger that the same phenomenon (inherited from their protolanguage) will be included several times as if from independent sources. (For instance,
a sample of languages for the study of tense-aspect systems based largely on Germanic
languages would overemphasize the incidence of the past/non-past tense opposition.)
Another bias to be avoided is geographic relatedness, since the study of language contact

235

236 Bernard Comrie

has shown that languages spoken in the same geographical area, in contact with one
another, will, with time, come to share various features. An often cited example is the
loss of the infinitive in languages of the Balkans (Rumanian, Bulgarian, Macedonian,
Modern Greek, Albanian, to some extent Serbo-Croatian), although recent work by
Matthew Dryer suggests that such areal grouping may in fact be much wider-ranging:
thus, northern Eurasia seems to be a linguistic area favoring AdjectiveNoun order,
whereas most of the rest of the world clearly favors NounAdjective order (Dryer
1989). A sample of languages biased towards northern Eurasia would therefore equally
be biased in favor of AdjectiveNoun order, although across the world as a whole this
is actually the less preferred order.
Early work within the Greenbergian tradition of language typology was primarily
empirically based, seeing as its initial task the establishment of a number of language
universals, on a solid empirical base, with explanation viewed as a subsequent stage
of investigation. In the meantime, developments within the field have led to a greater
emphasis on the search for explanatory principles underlying observed typological
generalizations. Some such explanations have been couched essentially in formal
terms, presupposing that formal simplicity is sufficient grounds for assuming that a
particular universal will hold; thus, Vennemann (1972) generalizes across several
of Greenbergs word order universals by positing two basic kinds of language from
the viewpoint of word order: operatoroperand (dependenthead) and operand
operator (headdependent), where head subsumes, for instance, the V of OV/VO,
the N (noun) of AN/NA (where A = adjective) and of GN/NG (where G = genitive),
the Ap of NPAp/ApNP (where Ap = adposition, i.e. preposition or postposition, and
NP=noun phrase), and dependent subsumes the other constituent of each of these
constructions. This generalization thus subsumes correlations among OV, AN, GN,
NPAp (all dependenthead), and among VO, NA, NG, ApNP (all headdependent).
As often in linguistics, there is a certain tension between more empirical and more
theoretical approaches: while Vennemanns approach does provide a much higher
level of generalization than does Greenbergs, the fit between predictions and observed
cross-linguistic distribution is much lower in the case of Vennemanns generalization
than in the case of Greenbergs more cautious statements.
More recently, attempts at explanation have gone beyond the purely formal realm,
attempting to draw parallels between the observed cross-linguistic distributions and
cognitive and pragmatic generalizations. For instance, Hawkins (1994) suggests that
many word order generalizations, including both universals and observed distributions of permitted alternative word orders, can be explained in terms of constraints
on language processing, such that those word orders are preferred that facilitate the
on-line comprehension of linguistic strings. Pragmatic explanations attempt to relate
the formal generalizations uncovered by empirically oriented typological research
to generalizations about language use. A simple illustration will refer back to the
implicational universal formulated above that if a language has first-person reflexive

Typology

pronouns it will also have third-person reflexive pronouns. The existence of a special
set of reflexive pronouns enables a distinction to be made with respect to the reference of pronouns, for instance in that in (1) he hit himself the two pronouns must be
coreferential whereas in (2) he hit him they must be non-coreferential. Differences of
referent are not normally crucial in the first (or second) person, since there is normally
only one first-person referent (the speaker) and only one second-person referent (the
addressee). However, in the third person, when potentially referring to any entity in
the universe of discourse other than speaker and addressee, the number of referents
is potentially large, and to be able to distinguish among referents is therefore often
crucial. It is thus not surprising that languages should have a greater tendency to make
overt differences of reference in the third person than elsewhere, a suggestion finding
further confirmation in other reference-tracking devices (e.g. obviation, logophorics),
which are also more likely to be found in the third person than in other persons.
While pragmatics can serve typology by providing explanatory models, as in the
example just suggested, the typological method can also serve (and has, indeed, served)
pragmatics through its emphasis on broadly-based cross-linguistic data. As is noted,
for instance, by Wierzbicka (1991:25 and passim), much work in pragmatics, rather
like early work in generative syntax, is based almost exclusively on observations relating to English, often leading to the formulation of generalizations that turn out to be
invalid when one turns to other cultures. The recent development of cross-cultural,
cross-linguistic studies of pragmatic phenomena has shown that there is in fact considerable variation among languages, variation that needs to be taken into account
in developing a general theory of pragmatics, as is illustrated by Wierzbicka (1991)
and several other works, including both works that emphasize universals, such as
Brown & Levinson (1987), and those that emphasize cross-linguistic diversity, such
as Verschueren (1985). As a preliminary to further development of this approach in
pragmatics, there is a need for the establishment of good data-bases, in terms both of
reliability and of cross-linguistic, cross-cultural representation, for the various subareas of pragmatics; a solid beginning in this direction has been made in only a few
areas, such as speech acts (including indirect speech acts) and linguistic politeness.
Just as the typological approach to grammar has given us a richer understanding of
human potential in the area of grammar, so the typological approach to pragmatics
promises to enrich our appreciation of human potential in the area of pragmatics.

References
Brown, P. & S.C. Levinson (1987). Politeness. Cambridge University Press.
Comrie, B. (1989). Language universals and syntactic typology. Blackwell/University of Chicago Press.
Croft, W. (2002). Typology and universals, 2nd edition. Cambridge University Press.
Dryer, M.S. (1989). Large linguistic areas and language sampling. Studies in Language 13: 257292.

237

238 Bernard Comrie

Greenberg, J.H. (1966). Some universals of grammar with particular reference to the order of
meaningful elements. In J.H. Greenberg (ed.) Universals of language: 73113. MIT Press.
Hawkins, J.A. (1994). A performance theory of order and constituency. Cambridge University Press.
Vennemann, T. (1972). Analogy in generative grammar, the origin of word order. In L. Heilmann (ed.)
Proceedings of the Eleventh International Congress of Linguists, vol. 2: 7983. Il Mulino.
Verschueren, J. (1985). What people say they do with words. Ablex.
Wierzbicka, A. (1991). Cross-cultural pragmatics. Mouton de Gruyter.

Variational pragmatics
Klaus P. Schneider

1. Introduction
Variational pragmatics is a field of inquiry at the interface of pragmatics and sociolinguistics. It can be defined as the study of intra-lingual pragmatic variation, i.e. of
pragmatic variation across varieties of the same language. More specifically, variational
pragmatics can be characterized as the intersection of pragmatics with dialectology.
In this context, dialectology is not understood in its narrow traditional sense as the
study of regional variation alone, but in a more comprehensive sense as the study of
both regional and social variation. In this sense, dialectology combines and integrates
the fields of dialect geography and urban dialectology and covers all areas of language
variation in sociolinguistics (cf., e.g. Schneider 2005b).
So far, the impact of such factors as social class, age and, especially, region on language use has been largely neglected in the field of pragmatics, particularly in empirical
speech act-based work. At the same time, the pragmatic domain of language has been
ignored almost completely in dialectology. Thus, variational pragmatics addresses two
complementary research gaps (cf. Schneider & Barron 2008a: 27 for details). Its aim
is, therefore, two-fold. On the one hand, it is aimed at introducing the examination of
regional and social variation in pragmatics research. On the other hand, it is aimed at
adding the pragmatic level to the other language levels overwhelmingly analysed in
dialectology, i.e. especially phonology, lexis and morphosyntax. In short, the overall
aims of variational pragmatics can be summarized as the dialectologisation of pragmatics and the pragmaticisation of dialectology (cf. Schlieben-Lange & Weydt 1978;
Schneider & Barron 2008a: 715).
The purpose of the present article is to outline the theoretical and methodo
logical foundations of variational pragmatics and to summarize its development to date.
Section2 focuses on crucial issues concerning pragmatics. In this section, the concept
of pragmatics underlying variational pragmatics is specified (2.1), the relationship
between variational pragmatics and related disciplines in pragmatics is described (2.2),
and five levels of pragmatic analysis are introduced (2.3). Section 3 deals with relevant
aspects of language variation. First, the concept of variation adopted in variational
pragmatics is explained (3.1). Then, pragmatic variation is discussed (3.2), and finally,
the question as to what constitutes a pragmatic variable is addressed (3.3). Section 4 is
concerned with methodological considerations. After outlining some methodological

240 Klaus P. Schneider

principles (4.1), the data types and methods of data collection employed in variational
pragmatics are characterized (4.5). Section 5 summarizes the history and development of variational pragmatics (5.1) and provides an overview of the most important
research trends (5.2). In the last section (Section6), areas in variational pragmatics are
identified which merit further study and open up perspectives for future research.

2. Pragmatic coordinates
2.1 The scope of pragmatics
Pragmatics has been defined in various different ways, ranging from the narrow
definition in the Anglo-American tradition to the very broad Continental European
tradition (cf., e.g. Huang 2010). The Anglo-American tradition, which is associated
with Gricean, neo-Gricean and post-Gricean approaches and concerns itself with such
phenomena as implicatures, explicatures, presuppositions and deixis, can, by and large,
be characterized as philosophical and deductive. Pragmatics is seen as an extension to
semantics (cf. also Huang 2007:4). In the Continental European tradition, by contrast,
pragmatics is a particular functional perspective on structural phenomena on all levels
of language (cf. also Verschueren 1999:7).
The notion of pragmatics so far advocated in variational pragmatics is not as
extreme as either of these two definitions, but marks an in-between position, which is,
however, closer to the Continental European understanding. In this regard, the phrase
adding the pragmatic level to the other language levels overwhelmingly analysed in
dialectology, which is used in Section1 to characterize one of the two over-arching
aims of variational pragmatics, is in fact misleading, as it suggests that pragmatics is
just another language level which is added to the levels of phonology, morphology,
syntax and semantics. It suggests, in other words, a position known as the component view (cf. Huang 2007: 4). The view which is adopted instead in variational
pragmatics can be called a complementary view. It is based on Leechs conceptualisation of grammar (i.e. the language system) and pragmatics (i.e. language use) as two
complementary and interacting domains (Leech 1983:4). These two domains interact
via semantics (as one level of the language system) (Leech 1983:12).
The standard textbook definition of the field of pragmatics as the study of language use is not sufficiently specific. What is crucial for variational pragmatics is
the literal meaning of the term pragmatics, which translates as the study of action,
where action is defined as intentional behaviour. Thus, the fundamental insight of
speech act theory that speaking is doing is of central importance. This does not mean,
however, that pragmatics can be reduced to the study of speech acts. At the same
time, there is no denying the fact that speech acts have played a decisive role in many

Variational pragmatics

subdisciplines of empirical pragmatics, notably those aimed at practical purposes such

as language teaching (i.e. contrastive, intercultural, cross-cultural and interlanguage
pragmatics; cf. Section2.2 below). Speech acts also play an important role in sociolinguistic approaches which integrate the pragmatic level (cf., e.g. Ervin-Tripp 1976;
Holmes 1986, 1990, 1995). Speech acts also occupy an important place in historical
pragmatics (cf., e.g. Jucker & Taavitsainen 2008) as well as in variational pragmatics
(cf. Section5.2 below).
Speech act theory and also early politeness theories (especially Brown & Levinson
1978, 1987; Leech 1983) have been criticized for, among other issues, their narrow focus
on speakers and individual utterances (cf., e.g. Watts 2003: 6369, 85116). Hence,
more recent pragmatic approaches have emphasized the role of hearers and their interpretations and a dialogic perspective. Accordingly, the focus in these approaches is on
longer stretches of discourse or on entire conversations and other speech events, and on
the interaction between the participants. Also, the notion of politeness (as the exclusive focus of early politeness theories) has been supplemented by notions of appropriateness and rudeness (or impoliteness) in a broader framework of relational work
(or face work) (cf., e.g. Locher & Watts 2005).
While for some scholars, in particular, but not exclusively, for many working in
the so-called Anglo-American tradition (cf. Huang 2007:4), these additional aspects
and more complex phenomena are outside the scope of pragmatics, but belong to
conversation analysis, discourse analysis or interactional linguistics, these aspects are
inside the scope of pragmatics as understood in variational pragmatics. All phenomena
studied constitute a range of different levels of analysis (cf. Section2.3 below).
2.2 The place in pragmatics
As the very name suggests, variational pragmatics is a branch of pragmatics. As such,
it is related to a range of other branches or subdisciplines of pragmatics which also
examine pragmatic differences. These are historical pragmatics, contrastive pragmatics, intercultural pragmatics, cross-cultural pragmatics, interlanguage pragmatics,
post-colonial pragmatics, and ethnopragmatics (cf. also Barron & Schneider 2009a:
425426). This list may not even be complete. The common denominator of all of
these subdisciplines is that they all work empirically, i.e. language data are not fabricated, but collected and analysed. In other words, they are all branches of empirical
pragmatics (cf. also Section4.1 below).
The label historical pragmatics is usually used to refer to two areas in which older
varieties of a language are examined and which are, strictly speaking, not identical
areas (cf., e.g. Jucker 1995). One is the area of diachronic pragmatics, which is concerned with pragmatic change in the course of the history of a language. The other area
is historical pragmatics proper, so to speak, i.e. the study of pragmatic phenomena

241

242 Klaus P. Schneider

(such as speech act realizations or conversational openings) in a particular historical

period of a language, e.g. in Early Modern English or Middle High German.
The term contrastive pragmatics pertains to the comparison of pragmatic pheno
mena across different languages (cf., e.g. Oleksy 1989). As a rule, only two languages
are involved, as contrastive pragmatics belongs to the general paradigm of contrastive analysis, in which two languages are contrasted mainly for practical reasons
(cf.,e.g. James 1980). As in the cases of, e.g. contrastive phonology and contrastive
grammar, the aim is to identify similarities and differences between two languages
for the purposes of foreign language teaching and foreign language learning. The two
languages contrasted are the native and the foreign language of the learners, and the
basic assumption underlying the comparison is that differences impede and similarities
facilitate the learning process, or, at least, that knowledge about differences helps to
understand errors made by the learners.
The terms intercultural pragmatics and cross-cultural pragmatics are often used
interchangeably (cf., e.g. Blum-Kulka et al. 1989a; Blum-Kulka et al. 2008). Crosscultural pragmatics is, in fact, similar to contrastive pragmatics. Cross-cultural pragmatics also compares two, sometimes more, languages, and, generally speaking, the
motivation behind analysing languages cross-culturally is practical application in
learning or training contexts. To pinpoint the difference between contrastive pragmatics
and cross-cultural pragmatics, it is useful to appeal to the distinction between pragmalinguistics and sociopragmatics as introduced by Thomas (1981) and popularized
by Leech (1983). In this view, pragmalinguistics deals with the linguistic devices available in a language to realize communicative functions, e.g. to perform a particular
speech act (as, for instance, how about-constructions for realizing a suggestion or an
indirect request), while sociopragmatics deals with the social situation in which the
communication takes place, addressing questions such as who uses which realization
of a given speech act to whom under which circumstances (cf. Leech 1983:1011).
Referring to this distinction, we can say that contrastive pragmatics is concerned with
pragmalinguistic aspects. Cross-cultural pragmatics, on the other hand, is concerned
with sociopragmatic aspects, thus not only focusing on language, but also integrating
a cultural dimension. The term intercultural pragmatics can be used to refer to this
same field of inquiry.
If, however, a distinction is made between cross-cultural pragmatics and intercultural pragmatics, then intercultural pragmatics deals with language use in communication between speakers who do not share the same native language. Two different
constellations are distinguished. One of these is native speaker non-native speaker
communication, i.e. a constellation in which the language used for communication is
the native language of one of the interlocutors, but a second or foreign language for the
other. The second type is lingua franca communication, i.e. a constellation in which

Variational pragmatics

the speakers do not share their native language either, but neither of the interlocutors
is a native speaker of the language used (cf., e.g. House 2003).
Interlanguage pragmatics concentrates first and foremost on the performance of
foreign language learners. This branch of pragmatics is based on Selinkers notion of
interlanguage (Selinker 1972). Interlanguage is the dynamic language variety used by
learners, which is heavily influenced by the learners native language in the beginning
stages, but becomes more and more similar to the target language in the course of the
learning process. Stages in this process are characterized by specific types of pragmatic
errors (cf., e.g. Trosborg 1995).
Post-colonial pragmatics does not deal with language performance by foreign language learners but by second language users, e.g. speakers of French in Cameroon or
English in India (cf. Janney 2009). Its aim is to analyse the use of the colonists language
in post-colonial times in those countries in which this language has been retained as
an official language, serving as a lingua franca in public life between speakers of sometimes several hundred different languages (an estimated 200 to 400 each for Nigeria
and India), and thus reducing or avoiding interethnic conflict. In such situations, there
is also interference from native language and native culture (cf., e.g. Sridhar 1991).
Ethnopragmatics is the term used for the type of analysis of intercultural communication developed by Anna Wierzbicka (cf., e.g. Goddard 2006). The aim of this type
of analysis is to compare cultures by using a semantic metalanguage, so-called Natural
Semantic Metalanguage (or, for short, NSM) as a tertium comparationis. This metalanguage, which is claimed to be universal, includes semantic primes (i.e. semantic
primitives) and also syntactic frames. Employing this metalanguage, so-called cultural scripts are formulated to capture culture-specific practices (cf., e.g. Goddard&
Wierzbicka 2007). Generally, cultures are compared whose members speak different languages, e.g. English and Polish (cf., e.g. Wierzbicka 1991). Recently, however,
national native-language varieties of English have also been compared, especially
British, American and Australian English (cf. Goddard 2009). Predominantly, fictional
material and indirect evidence have been used in ethnopragmatics, which could therefore be regarded as the least empirical of the pragmatic disciplines discussed in this
section. However, these other disciplines have been strongly inspired by Wierzbickas
work, notably her 1985 article, in which she attacks universalist claims and (tacit)
ethnocentrism in early speech act theory, politeness research and conversation analysis
and makes a plea for intercultural studies in pragmatics (cf. Wierzbicka 1985, also
Blum-Kulka et al. 1989a).
Surprising as it may seem, it can be argued that variational pragmatics is most
similar to historical pragmatics, because both disciplines study pragmatic variation in
one and the same language, while in all other disciplines more than one language is
involved. The difference between historical and variational pragmatics is that variational

243

244 Klaus P. Schneider

pragmatics examines variation in space, i.e. more specifically (and more metaphorically)
in regional and social space, and historical pragmatics examines variation in time. This
holds true for diachronic pragmatics in particular, whereas historical pragmatics in the
narrow sense of the term is as much a branch of synchronic linguistics as is variational
pragmatics. Therefore, in addition to establishing the nature of pragmatic phenomena
in an earlier historical period, it is perfectly conceivable to also study pragmatic variation across social groups in the same period of time, provided suitable data material is
available (cf. Jucker & Taavitsainen forthcoming).
Irrespective of how many languages are involved just one, or more than one all
disciplines discussed here study pragmatic variation across cultures. The differences
found across languages, or across varieties or historical stages of the same language
can be attributed to diverging cultural norms and values. Hence, it would seem befitting to use the term cross-cultural pragmatics as a cover terms for all disciplines
discussed. This is, however, problematic considering the history and conventional
use of this term, which is more restricted in scope (as is, e.g. the term American
English, which traditionally refers only to English as it is spoken in the United States,
but not to other American varieties such as Canadian English or Caribbean English).
Therefore, cross-cultural pragmatics is used in its conventional narrow sense in the
present context.
2.3 Levels of pragmatic analysis
Variational pragmatics distinguishes five levels of pragmatic analysis. These are the
formal level, the actional level, the interactional level, the topic level, and the organisational level (cf. Schneider & Barron 2008a: 1921). This framework may well be called
eclectic, as it aims at integrating different approaches and traditions. No claims are
made that this list of five levels is exhaustive.
The formal level of pragmatic analysis takes as its starting point a linguistic form
and aims at determining the various communicative functions this form may have in
discourse. The analytic approach taken is, in other words, form-to-function mapping.
Typical examples of such forms are well, you know, I see, etc. This level serves to integrate work on discourse markers in particular.
The actional level is the level of linguistic action, i.e. of speech acts. On this level,
the starting point is a communicative function, i.e. an illocution such as request,
promise or apology, and the aim is to identify the linguistic realizations with which
this illocutionary act can be performed. Here the approach is not form-to-function
mapping, as on the formal level, but the opposite, i.e. function-to-form mapping. This
level is based on insights from philosophical speech act theory as well as from linguistic speech act analysis. In the latter tradition, a number of distinctions can be made.
Afirst distinction is that between pragmalinguistics and sociopragmatics, as discussed

Variational pragmatics

in Section2.2 above. A second distinction is that between conventions of means and

conventions of form (cf. Clark 1979). Conventions of means concern realization
strategies, whereas conventions of form concern the concrete linguistic choices made
when applying these strategies. One such realization strategy is the so-called suggestory formula which can be used to indirectly perform a request (cf. Blum-Kulka etal.
1989a). The term is motivated by the fact that the conventions of form are primarily
used to directly perform a suggestion. These conventions of form include, for instance,
syntactic constructions beginning with How about (doing X) and Why not (do X).
Finally, a distinction can be made between speech act realization and speech act modification (cf. Blum-Kulka et al. 1989a). Research in the area of speech act realization is
aimed at identifying the inventory of conventions of means and the corresponding
conventions of form available for performing a given illocutionary act. Research in the
area of speech act modification, on the other hand, examines linguistic devices which
can be used to modify realization strategies. For example, an imperative construction
employed to realize a request (a strategy known as mood derivable; Blum-Kulka et al.
1989a) can be modified by adding please, or perhaps, or a question tag (e.g.will you?).
These modificational devices are not specific to any particular strategy. The modifiers
please or perhaps, for instance, can also be used in other strategies (e.g. Can you open
the window, please?). As can be seen from this discussion, phenomena such as directness and indirectness are also addressed on this level, as are issues of politeness relating
to individual speech acts (in the frameworks of early politeness theories).
The interactive level deals with units beyond individual utterances or speech acts.
The units examined are dialogic and range from paired utterances through longer
stretches of discourse to complete conversations or other types of talk. These units
include, among others, adjacency pairs (alternatively referred to as interactional
exchanges), speech act sequences, conversational openings and closings in short, local
and global patterns of sequential organisation in spoken discourse (cf., e.g. Edmondson
1981; McLaughlin 1984). Further phenomena which can be studied on this level are the
construction of identities, the negotiation of politeness and rudeness and other aspects
of relational work (cf., e.g. Locher 2008).
On the topic level, issues of topic selection and topic management are addressed.
These concern the way in which topics are introduced, maintained, developed, changed
and terminated (cf., e.g. Schneider 1988:8197). Also included is the question which
topics may be selected under which circumstances, e.g. which topics may be chosen
in conversations with strangers or with superiors at work, or which topics are taboo in
small talk (cf., e.g. Wolfram & Schilling-Estes 2006:98). Prototypically, topics analysed
on this level are macropropositions, i.e. they relate to the contents of discourse units
consisting of several turns at talk. However, topic analysis can also refer to individual
utterances, or, more precisely, the proposition of an individual speech act. Compliment
topics are a case in point, i.e. the item which the positive evaluation expressed in the

245

246 Klaus P. Schneider

compliment refers to (cf., e.g. Holmes 1995). A relevant question in this regard is who
may compliment whom on what. In short, on this level discourse content is examined.
The shaping of this analytic level has been largely inspired by discourse analysis, notably
psycholinguistically oriented discourse analysis (cf., e.g. Brown & Yule 1983).
The organisational level serves to integrate findings from ethnomethodology and
conversation analysis. First and foremost, it deals with the mechanisms of turn-taking,
i.e. how speakers take turns at talk and how the floor is distributed. Further issues
such as interruptions and silence in discourse are directly related to these questions
(cf., e.g. Sacks et al. 1974, also Wolfram & Schilling-Estes 2006:98100).
Distinguishing different levels of pragmatic analysis is necessary, because the
analysis must not be limited to an examination of speech acts alone. Speech act analysis
is just one of the five levels distinguished in the present framework, i.e. the actional
level. Distinguishing different levels of pragmatic analysis is also necessary because,
in the complementary view, in which language system (grammar) and language
use (pragmatics) are conceptualised as complementary domains of equal status
(cf.Section2.1 above), it seems desirable to subdivide language use by analogy to
mutatis mutandis the subdivisions of the language system, i.e. the levels of phonology,
morphology, syntax and semantics.
Finally, distinguishing different levels of pragmatic analysis is necessary to come
to grips with the complexities of verbal communication. Needless to say, the five levels
of analysis specified in the present framework of variational pragmatics are only analytically separated. It is also conceivable to distinguish different levels. Similar frameworks
(or discourse models) have been proposed by, e.g. Schiffrin (1987), Schneider (1988)
and Jucker (2008).

3. Variational coordinates
3.1 Variation and identity
The distinction of five levels of pragmatic analysis forms one part of the analytical
framework of variational pragmatics. The other major part is the distinction of five
types of language variation. These are regional variation, socioeconomic variation,
ethnic variation, gender variation, and age variation (cf. Schneider & Barron 2008a:
1619 for further details). In other words, one basic assumption in variational pragmatics is that the factors region, social class, ethnicity, gender and age influence language use in (inter)action in systematic ways. These factors are generally referred to as
sociological, sociolinguistic, or macro-social factors. The term macro-social factors is
preferred in the present framework.

Variational pragmatics

Regional variation has been studied, more or less exclusively, in traditional dialectology, notably in dialect geography. Originally, researchers concentrated on non-mobile
old rural males (known as NORMs) as the most conservative group of speakers in a
population, who are using the highest proportion of typical features of regional speech
in a dialect area. The results of this type of research have been documented in dialect
dictionaries and linguistic atlases (cf., e.g. Schneider 2005b).
The other four types of variation (socioeconomic, ethnic, gender and age variation)
have been studied in sociolinguistics, or, more specifically, in urban dialectology. As
the latter term indicates, here the focus has been not on rural areas, but on urban
centres, i.e. cities such as New York or Belfast (cf., e.g. Labov 1966b; L. Milroy 1980;
J.Milroy 1981). Urban dialectology has not, as a rule, concentrated on only one particular group of speakers, but on carefully stratified populations. This line of research
is associated with William Labovs name and with the methodological principles and
innovations Labov has introduced in sociolinguistics. This paradigm is known as the
variationist (not: variational) paradigm.
Present-day dialectology considers both regional variation, as studied in dialect
geography, and the types of variation examined in urban dialectology. This holds true
in particular for dialectology in the United States of America (cf., e.g. Wolfram &
Schilling-Estes 2006). However, in other countries, such as Germany dialectology
and sociolinguistics are still seen as two distinct disciplines investigating different
types of language variation. In order to avoid confusion, that kind of dialectology
which deals with all types of variation can be referred to as integrative dialectology
(cf. Schneider 2005b).
Originally, the term social variation was used to refer to the impact of social class
on language exclusively, for which socioeconomic variation is now used, to avoid
ambiguity. Today, social variation is a cover term for socioeconomic, ethnic, gender
and age variation, which is usually contrasted with regional variation. In variational
pragmatics, however, regional variation is also subsumed under social variation,
since what is of primary interest is not geographical fact, but regional affiliation and
identity. Thus, regional variation is treated in the same way as, e.g. gender variation.
Gender studies emphasize that social sciences (including sociolinguistics) and cultural studies are not interested in sex, i.e. the dichotomous biological category, but
in gender, i.e.the social construction of a wide range of gender identities. Likewise,
the study of language variation is not interested in race, but in ethnicity, and similar
claims are made about age (chronological versus psychological age) and social class
(job or income versus affiliation). So, all five factors examined in variational pragmatics
are ultimately seen not as crude facts, but as social constructs. In all cases, we are dealing
with aspects of identity, at least theoretically. From a practical, research-strategic and
methodological point of view, however, it is not a trivial task to gain access to identities

247

248 Klaus P. Schneider

and to adequately analyse them in empirical work. Cheshire (2002:423425) argues

that gender variation can only be studied adequately in long-term projects and by
employing ethnographic methods such as participant observation. Against this background, Cheshire proposes to at least initially concentrate on the more obvious sex of
speakers, as most researchers in sociolinguistics actually analyse sex differences rather
than gender differences, despite the fact that some of these researchers claim to be
focusing on gender, if only because they think speaking of speaker sex in this context
is not politically correct. By concentrating on sex, Cheshire further points out, it is
easier to compare sociolinguistic studies of this type of variation and their results. In
variational pragmatics, it is assumed that the same applies to investigations into the
other four types of variation as well.
Even though the concept of social class has played a vital role in the development of
sociolinguistics, it has been criticized for being heavily undertheorized (cf.,e.g. Downes
1998:228; Ash 2002:419420). It seems that distinctions such as working class versus
middle class, or lower middle class versus upper middle class, are not informed by
present-day sociological theory. Therefore, approaches focusing on social networks
rather than on social class have been advocated as an alternative (cf., e.g. Downes
1998). Other scholars have argued that education and/or income may be more relevant
categories than social class (cf., e.g. Plevoets et al. 2008, who distinguish between
educational level and occupational level). It is possible that all three factors, i.e. social
class, income and education, have to be taken into account to arrive at an adequate
analysis, and maybe also social networks. In any case, no claims are made in variational pragmatics that the five macrosocial factors considered presently form a closed
set. It is perfectly conceivable to add further factors to this list. For instance, religious
affiliation, which seems to play a decisive role in some societies, is a candidate for
inclusion. In the long run, it may be inevitable to work across disciplinary boundaries
with scholars from sociology, ethnology and anthropology.
Region, which has been a central notion in traditional dialectology, is understood
in variational pragmatics as an umbrella term for a hierarchy of spatial entities. In
dialect geography, regional variation has been studied across regions in the same
country. Fairly independent of this tradition, a branch of research has been established in which national varieties of languages are examined, specifically national
varieties of so-called pluricentric languages, i.e. languages spoken in different countries around the world (cf., e.g. Clyne 1992). Pluricentric languages are, for instance,
English (as spoken in the U.K., the U.S., Ireland, Australia, and so forth) and Spanish
(as spoken in Spain, Mexico, Peru, Uruguay, and so forth). Generally, the focus is on
first language varieties (or inner-circle varieties, in Kachrus model; Kachru 1985),
often second language varieties (outer-circle varieties) are also dealt with (as well as
Pidgins and Creoles based on the respective language) (cf., e.g. Trudgill & Hannah 2002);
foreign language varieties (expanding-circle varieties) are, however, usually excluded

Variational pragmatics 249

(but cf. Melchers & Shaw 2003). Sociolinguistics, on the other hand, has dealt with cities
such as New York City or Norwich, or with areas in cities, e.g. Catholic or Protestant
working class and middle class areas in Belfast (cf., e.g. Milroy 1998). Variational pragmatics integrates all of these perspectives in the study of regional variation and distinguishes between national, sub-national (i.e. regional in a dialect geographic sense),
local, and sub-local varieties. A further distinction which cuts across these regional
divisions is the rural urban divide. It seems that language use and communicative
behaviour in rural areas is more similar to each other across regions than to urban
areas or cities in the same region. This particular type of regional variation, which has
so far not received much attention, merits further investigation.
3.2 Pragmatic variation
Scholars working in the Anglo-American tradition of pragmatics, including speech
act philosophy, (neo-/post-) Gricean approaches and first-generation politeness theories, are generally not interested in variational issues. They are primarily interested
in pragmatic universals, i.e. the fundamental workings of human communication
(cf.Section2.1 above). This tradition has been accused of ethnocentrism and criticized for mistaking language-use conventions in the English-speaking world for
pragmatic universals (cf., e.g. Wierzbicka 1985, also Hill et al. 1986, and Section2.2
above). Similarly, conversation analysis, at least the original version of CA developed
in ethnomethodological sociology, has been interested in the general workings of,
e.g. repair and turn-taking (cf., e.g. Sacks et al. 1974), but not in regional and social
variation (cf, e.g. Schegloff & Sacks 1973:291). There are, however, empirical studies
adopting CA methods in cross-cultural research (cf., e.g. Golato 2002).
Most branches of empirical pragmatics, on the other hand, are definitely focused
on variation. This certainly applies at least to both contrastive and cross-cultural
pragmatics. These disciplines are, however, dealing with pragmatic differences across
languages, and not across varieties of the same language. In this particular context,
however, languages are conceptualised as homogeneous wholes, as if all speakers of a
language made use of this language in communication in uniform ways, which is an
assumption that has been criticized as reductionist, e.g. by Kasper (1995). This type of
pragmatic variation can be termed inter-lingual variation.
Inter-lingual variation contrasts with intra-lingual variation, i.e. variation not
across languages, but in a given language. Two types of intra-lingual variation can
be distinguished. These are micro-social and macro-social variation. Micro-social
variation concerns speaker constellations and is caused by such factors as power and
distance. There are two basic types of power constellations, namely symmetrical communication, i.e. communication between speakers of equal social status such as workmates, friends or neighbours, and asymmetrical communication, i.e.communication

250 Klaus P. Schneider

between speakers of different social status, as between employer and employee or

teacher and student. In empirical studies, it has to be specified whether the speaker is
in a superior or in an inferior position vis--vis the addressee, or whether the interactants are of equal status. Social distance, on the other hand, can be measured on a
continuum. Yet, as a rule, many empirical studies distinguish solely between +D,
i.e. maximal distance, as in stranger-stranger constellations, and D, i.e. minimal
distance, as between (close) friends, family members and lovers. It has been shown,
however, that the in-between constellations are of particular interest, since constellations characterized by middle distance (for instance between acquaintances)
require the highest degree of investment in, e.g. politeness (cf. Wolfsons bulge theory;
Wolfson 1988). These two factors in particular, i.e. power and distance, have been
systematically analysed and thus played an important role in cross-cultural pragmatics (cf.,e.g. Blum-Kulka et al. 1989a), and also in Brown & Levinsons politeness
theory (1978, 1987).
What has, by contrast, been much neglected is macro-social variation, i.e. the
impact of such factors as gender, age, ethnicity, social class and region (cf. Kasper
1995, also Wierzbicka 1985). It is a major aim of variational pragmatics to address
exactly this research gap (cf. the dialectologization of pragmatics, Section1 above).
In conclusion, then, variational pragmatics can be defined as the study of intra-lingual
macro-social variation.
The research agenda for variational pragmatics comprises several steps. A first
step consists in analysing the impact of each of the macro-social factors individually.
In this regard, a basic requirement is that all other factors are controlled and kept
invariable. As all speakers are at the same time either male or female and belong to a
particular age group, ethnicity, social class and region (place, nation, etc.), a second
step involves investigating the interplay of the different macro-social factors. It is, as
yet, an open question whether each factor configuration is associated with specific
language-use conventions, or whether there are typical patterns of communicative
behaviour which are shared e.g. by women across age groups or ethnic communities,
and so on. As speakers interact in a range of different constellations, a third step in the
analysis deals with the interplay of macro-social and micro-social factors. At present,
it is not sufficiently clear how further relevant concepts can be fruitfully integrated into
the current framework of variational pragmatics. Such concepts include genres, registers, styles, activity types, communities of practice, and relational work (cf. Schneider
2007; Haugh & Schneider in prep.).
3.3 What is a pragmatic variable?
The last theoretical issue to be discussed here concerns the concept of the linguistic variable relevant to work in variational pragmatics. The concept of the linguistic

Variational pragmatics

variable was originally developed by Labov (1966a) for application in the analysis of
phonological variation. This concept has been discussed extensively ever since. Inparticular, the assumption has been challenged that this concept can be applied in the
sociolinguistic analysis of variation on language levels other than phonology. The
crucial issue in this discussion is the notion of sameness, which is defined in truthconditional terms for the purposes of phonological analysis. It is not clear, however,
how sameness can be defined for the analysis of syntax, for example (cf., e.g. Lavandera
1978). Likewise, it is not immediately clear how sameness is to be defined for the purposes of analysing pragmatic variation.
It has been claimed that variational pragmatics is firmly rooted in the Labovian
paradigm (cf. Terkourafi forthcoming). This is understandable, as variational pragmatics is first and foremost interested in finding out about the systematic influence of
macro-social factors on language use. Moreover, it is true that variational pragmatics
is aimed at determining pragmatic variables and their variants (cf., e.g. Schneider&
Barron 2008a: 11, 22) variables and variants are technical terms widely used in
sociolinguistic research la Labov. In this respect, variational pragmatics is obviously related to the fundamentals of variationist sociolinguistics. At the same time,
however, variational pragmatics is, at least theoretically, open to constructivist ideas
(cf.Section3.1 above). It is not sheer coincidence that variational pragmatics is called
variational pragmatics and not variationist pragmatics. The name actually chosen
points to the fact that variational pragmatics is not considered only as an extension of or addition to variationist sociolinguistics. This means, among other things
(cf.Section4 below), that the terms variable and variant are used and defined in
variational pragmatics more loosely than in variationist work.
The use of variable and variant in variational pragmatics may, in fact, seem
naive. In this framework, a pragmatic variable is that which varies on the pragmatic
level across varieties of the same language, and the respective variants are the options
available of which different social groups of speakers make use. For instance, on the
actional level of analysis, a variable is, in the most general terms, the speech act. It
is assumed that speech acts exist in all speech communities or cultures. It may well
be the case that not all speech acts exist in all speech communities or cultures, but
there is sufficient evidence to suggest that such speech acts as requests exist at least
in many or most speech communities. For each speech act, there is a range of different realisation strategies, even for speech acts with the same illocution and the same
proposition in the same type of situation. In other words, sameness can in this case be
defined in functional (i.e. illocutionary), propositional and situational terms. So what
varies is, in fact, not the illocution, but the realisation strategy (or, more precisely, the
conventions of means and the conventions of form), and each realisation is a variant
of this variable. On the interactional level of analysis, to provide a further example, it
has, for instance, been found that different social groups open a casual conversation

251

252 Klaus P. Schneider

with a complete stranger at a social gathering in different ways, i.e. by using different
speech acts (e.g. a greeting such as Hi, or a remark about the occasion such as Great
party, isnt it?; cf. Schneider 2008a). In this case, sameness is defined positionally (or
via the communicative task at hand, i.e. opening a conversation in this type of situation). Finally, on the formal level of pragmatic analysis, sameness is identity of form,
e.g. the discourse marker well. What varies are the different communicative functions
this particular marker may serve, and each function identified is a variant of this
variable (cf. Jucker & Taavitsainen, forthcoming, for a similar view, and Terkourafi,
forthcoming, for a diverging position).

4. Methodological concerns
4.1 Some basic principles
Variational pragmatics is not a type of armchair linguistics (cf., e.g. Jucker 2009),
i.e.it does not use so-called introspective data such as episodical, intuitive or fabricated examples, which are commonly analysed in the Anglo-American tradition of
pragmatics (true introspection as used in psycholinguistic research is, however, something quite different, viz. a type of experimental method; cf., e.g. Faerch & Kasper
1987). Work in variational pragmatics, by contrast, is always empirical. This is the first
and most basic methodological principle of variational pragmatics.
A second methodological principle is the principle of contrastivity. Many socio
linguistic studies as well as many pragmatic studies focus on one variety alone. Examples are studies on youth language or studies of Irish English. While it is certainly
true that any such study can describe features of the variety under inspection, there
is no guarantee that these features are specific to this variety (cf. Barron & Schneider
2005a: 1112). Whether or not features are variety-preferential or variety-exclusive
can only be established by contrasting the variety focused on with data from another
variety. In the case of youth language, features specific to this variety can only be
determined by contrasting the speech of adolescents with the speech of adults, for
example. Likewise, Irish English data must be contrasted to data from, e.g. English
English, American English or New Zealand English. Hence, a contrastivity principle
of variational pragmatics has been posited, which is formulated as follows (Barron &
Schneider 2009a: 429): Linguistic features can be considered variety-specific only if
the variety under study is contrasted with at least one other variety of the same kind
and of the same language.
Furthermore, data sets to be contrasted have to be comparable. This is a third
methodological principle of variational pragmatics. There is no point in contrasting
speech by, e.g, old American working class males and young Australian middle class

Variational pragmatics

females, to give a crude example. Ideally, the data sets contrasted are produced by
speakers of the same sex, age group, social class, etc. These three principles, i.e. the
principles of empiricity, contrastivity and comparability, are the sine qua non of the
methodology of variational pragmatics.
4.2 Data
Concerning data types and methods of data collection, variational pragmatics is
not dogmatic (unlike other disciplines in pragmatics such as conversation analysis,
which permits no other material than naturally occurring conversation). In general,
it is firmly believed that no one method is the only right method. Each method
has advantages as well as disadvantages, and the choice of method ultimately depends
on the research questions to be answered (cf., e.g. Golato 2003; Kasper 2000, 2008;
Jucker 2009). Accordingly, a wide range of different data types and methods of data
collection have been used in variational pragmatics. Data types include naturallyoccurring discourse as well as experimental data. Naturally-occurring discourse has
been extracted from well-known and widely available large electronic corpora or from
small self-recorded data sets. As a rule, these corpora provide audio-recorded material
(usually only transcripts, sometimes also sound files), but no video-recordings. This
may change in the near future (cf., e.g. Clancy forthcoming). Large corpora of English
which have been used are, e.g. the British National Corpus, the Wellington Corpus of
Spoken New Zealand English, and the British and the Irish components of the International Corpus of English. Smaller corpora which have been analysed are the Cambridge
and Nottingham Corpus of Discourse in English (CANCODE) and the Limerick Corpus of Irish English (LCIE) (cf. Section5.2 below for examples). For the purposes of
variational pragmatics, it is essential that demographic information about the speakers, i.e. the macro-social participant features, are known. Not many of the best-known
corpora provide this kind of information. The Santa Barbara Corpus is a relatively rare
exception. In other cases, demographic information is incomplete, unreliable or not
available at all. As regards contrastivity and comparability, the most useful corpora are
those which are compiled and structured in the same way. This applies, e.g. to the various
components of ICE, and also to CANCODE and LCIE.
Experimental data have been elicited by employing such methods as role play and
different types of written production questionnaires (including discourse completion tasks and also dialogue production tasks) (cf. Section5.2 below for examples).
While experimental data in general and written questionnaires in particular are often
frowned upon in empirical pragmatics research, these types of data are especially
suitable for research in variational pragmatics, as they allow for maximal control of
macro-social factors. This is an aspect not to be underestimated. Furthermore, it has
been argued that data elicited by employing production questionnaires do not represent

253

254 Klaus P. Schneider

what people actually say, but what they would or should say and what is generally
expected in a given type of social situation. In other words, questionnaire data reflect
the respective norms of the social groups the informants belong to or affiliate with
(cf.,e.g. Schneider 2010).
The least which can be said is that it is a recommended strategy to start with
experimental data and then, in a second step, to compare the findings arrived at with
material from corpora of naturally occurring discourse. It is perfectly conceivable that,
in a further step, the results are tested, supplemented or elaborated in ethnographic
research to arrive at a more fine-grained analysis and a more comprehensive, differentiated and, thus, adequate picture, which may meet the standards of constructivist
approaches (cf. Section3.1 above).

5. Development
In the present section, an outline is given of the short history, and pre-history, of variational pragmatics (5.1) and an overview is provided of the work done in this field to
date (5.2). Specifically, the question is addressed how much attention the individual
levels of pragmatic analysis and the different macro-social factors have attracted
which are specified in the current framework of variational pragmatics (cf.Sections2.3
and 3.1 above).
5.1 History
Variational pragmatics has its roots in the 1990s (cf. also Barron & Schneider 2009a:
432434). The 1980s saw the beginnings of cross-cultural pragmatics. In the eighties, a large number of studies appeared that examined empirically pragmatic variation
between different languages. Interestingly, in the best known project at the time, i.e.
the Cross-Cultural Speech Act Realization Project (CCSARP), comparable data were
collected not only from different languages, but also from national varieties of the
same languages (cf. Blum-Kulka et al. 1989b). Cross-varietal comparison was, however, never explicitly made or published.
Then, in the 1990s, before the term variational pragmatics was even coined
and used, individual papers investigating pragmatic differences between national
varieties of the same language were published. Muhr (1994), comparing apologies in
the varieties of German spoken in Germany and in Austria, and Schneider (1999),
contrasting compliment responses in Irish English and American English, are early
examples. Also, a proposal was made to study pragmatic variation across the many
national varieties of Spanish (Placencia 1994). At about the same time, sociolinguistic
research was conducted on pragmatic differences between male and female language

Variational pragmatics

use (cf.,e.g. Holmes 1995). Integrating the study of pragmatic variation on the subnational level was considered a few years later (cf. Wolfram & Schilling-Estes 1998;
Schneider 2001), an area which had, in fact, been identified as a desideratum a long
time ago (cf. Schlieben-Lange & Weydt 1978), but no empirical work had been carried
out in the meantime (Schneider & Barron 2008a: 813).
It was in 2005 that variational pragmatics was officially introduced. At the 9th
International Pragmatics Conference (IPC 9) held in Riva del Garda, Italy, Barron and
Schneider organised a panel called Variational pragmatics: Cross-cultural approaches,
and they opened it with a programmatic paper on Variational pragmatics: Contours
of a new discipline. All other panel contributions dealt with pragmatic variation across
national and subnational varieties of English, Spanish or German, respectively.
Also in 2005, the first publication appeared which had variational pragmatics
in its title, i.e. the article Variational pragmatics in the foreign language classroom
(Barron 2005b). In this article, a plea is made for including speech act differences
between national varieties of the target language in the teaching of foreign languages.
That same year, a collection of papers on The Pragmatics of Irish English was published,
edited by Barron & Schneider (2005b). This collection was the first book devoted
exclusively to the study of a wide range of pragmatic phenomena in one particular
national variety of the English language. In their introductory chapter, the editors
already point out the possibility or, indeed, the necessity of variational pragmatics
(Barron & Schneider 2005a: 1113), but only three of the contributions to this book
are explicitly contrastive, i.e. comparing Irish English to the variety of English spoken
in England (Barron 2005a; Kallen 2005) and also to the variety spoken in the United
States of America (Schneider 2005a).
Variational pragmatics was further developed in two additional panels convened
by Barron and Schneider. The first of these dealt with Variational pragmatics: A focus
on region, age and gender and was organized at the 10th International Pragmatics
Conference (IPC 10) held in Gteborg in 2007. As the title suggests, this panel was
focused on the interaction of different macro-social factors. The second of these two
panels, on the other hand, dealt with the interaction of macro-social and microsocial factors (cf. Section3.2 above), as indicated in its title, Pragmatic variation: The
interplay of micro-social and macro-social factors. This workshop was held at the Sociolinguistics Symposium 17 (SS 17) in Amsterdam in 2008.
Later in 2008, a book was published with the title Variational Pragmatics: A
focus on regional varieties in pluricentric languages (Schneider & Barron 2008b). Asits
subtitle shows, this book concentrates on only one type of macro-social variation,
viz. regional variation (in one case on the subnational level, in all other cases on the
national level). However, the introductory chapter, authored by the two editors, provides a programmatic introduction to the entire field of variational pragmatics, specifying the research gaps which are addressed and developing in detail the framework

255

256 Klaus P. Schneider

summarized above in Sections2.3 and 3.1 (Schneider & Barron 2008a). The pluricentric
languages covered in the volume are English, Spanish, German, Dutch and French
(cf.also Section5.2 below).
Finally, in 2009, a special issue of the journal Intercultural Pragmatics was devoted
to variational pragmatics (Barron & Schneider 2009b). In the introductory chapter, the
editors broaden the discussion of theoretical and methodological issues in variational
pragmatics and set the scene for the six papers which analyse the interaction of macrosocial factors, specifically of regional, socioeconomic, gender and/or age variation, in
varieties of English and Spanish (Barron & Schneider 2009a). In at least two of the
papers, variety-specific (or even variety-exclusive) choices made to realize particular
communicative functions in discourse are explicitly regarded as displays of national
or local identities, respectively (OKeeffe & Amador Moreno 2009; Pichler 2009), thus
clearly extending the analysis beyond the statistical correlations of the variationist paradigm (cf. Section3.1 above). In a further contribution to this special issue, Macaulay
(2009) discusses discourse features employed by working-class adolescents as displays
of identity of members of a specific age and social class community.
5.2 Research trends
The present subsection provides a brief overview of the areas covered by research
in variational pragmatics carried out to date. Given its short history (cf. Section5.1
above), it is not surprising that the five levels of pragmatic analysis and the five macrosocial factors currently distinguished in variational pragmatics (cf. Sections2.3 and
3.1 above) have not all received the same amount of attention.
Currently, work in variational pragmatics can be characterized as follows. Prototypical studies compare the use of one speech act across two national native speaker
varieties of a pluricentric language. For instance, Barrons article on requests in
English English and Irish English (Barron 2008), and Garcas article on invitations
in Argentinean Spanish and Venezuelan Spanish (Garca 2008) are representative
examples of this type.
English, with about a dozen national native speaker varieties around the world, and
Spanish, with more than twenty national varieties spoken natively in Europe and the
Americas, are the two pluricentric languages which have been examined extensively in
variational pragmatics (cf., e.g. Schneider & Barron 2008a & Flix-Brasdefer 2009 for
overviews). In fact, the investigation of pragmatic variation across national varieties
of Spanish, sometimes labelled as sociopragmatics, can, in its origins, be considered
a parallel development to the development of variational pragmatics, a term initially
applied only to the investigation of pragmatic variation across varieties of English
(cf.Section5.1 above). Other pluricentric languages examined in variational pragmatics include German (cf., e.g. Muhr 2008), Dutch (cf., e.g. Plevoets et al. 2008),

Variational pragmatics

and French (cf., e.g. Schlmberger 2008). So far, work in variational pragmatics has
focused on Indo-European languages. It would seem desirable to also include other
languages, especially such pluricentric languages as, e.g. Arabic or Swahili.
While many national varieties of English and Spanish have been examined, not all
varieties of these two languages have been covered. National varieties of English which
have been compared are British English (or, more specifically, English English), Irish
English, American English, and New Zealand English (cf., e.g. Jautz 2008 contrasting
British and New Zealand English). Australian English, on the other hand, and especially Canadian English have so far been largely neglected.
The varieties of Spanish which have been examined in variational pragmatics
(and in earlier intra-lingual comparison in sociopragmatics; cf. Section5.1 above)
include Peninsular (i.e. European), Argentinean, Chilean, Colombian, Costa Rican,
Dominican, Ecuadorian, Mexican, Peruvian, Uruguayan and Venezuelan Spanish
(cf.,e.g. Marquez Reiter 2002 contrasting Peninsular and Uruguayan Spanish). Not
yet examined are the varieties of El Salvador, Guatemala, Panama and Paraguay
(cf.Flix-Brasdefer 2009:506507).
As has been mentioned above, in most cases two varieties are compared. The comparison of three or more varieties is rare. Examples of this rare type include Schneider
(2005a and 2008a), comparing three varieties of English (viz. English English, Irish
English, and American English), and Flix-Brasdefer (2009) and Garca (2009), each
comparing three varieties of Spanish (viz. Mexican, Costa Rican and Dominican
Spanish, and Argentinean, Peruvian and Venezuelan Spanish, respectively).
By comparison, studies of regional variation on the subnational level are rare.
Examples include Barron (2009) and Placencia (2008). In Barrons study, U.S. American
apologies are analysed. Specifically, the author compares apologies among male speakers
in Virginia and in Tennessee. Placencia, on the other hand, examines pragmatic differences in Ecuador by contrasting Quito, the capital city in the Andean mountains, and
Manta, the harbour city on the Pacific ocean. Her analysis concentrates on requests
in corner shop interactions. While Pichler (2009) and Macaulay (2009) also focus on
individual cities viz. Berwick-upon-Tweed in the north of England and Glasgow,
respectively they do not contrast their findings with findings from other cities.
Macaulay, however, considers gender and age variation and also language change on
the pragmatic level, whereas Pichler examines doing regional identity by speakers
employing local variants of discourse markers as opposed to standard equivalents of
these markers.
Gender differences or, more appropriately, sex differences (cf. Section3.1 above),
in the pragmatic domain were investigated in sociolinguistics before variational pragmatics was conceived of as a field of inquiry. Holmess work on apologies and on compliments in New Zealand English is a case in point (cf., e.g. Holmes 1986 and 1990;
cf. also, e.g. Goodwin 1990 on directives among black girls and among black boys in

257

258 Klaus P. Schneider

the USA). In the explicit framework of variational pragmatics, sex differences have
been analysed first and foremost in its interaction with the factor age. Macaulay (2009)
has already been mentioned in the preceding paragraph; a further example is Farr &
Murphy (2009). While Macaulay examines Scottish male and female adolescents in
different age groups, Farr & Murphy examine different groups of Irish male and female
adults. The interplay of sex and age variation is also analysed in Murphy (2010).
Episodic evidence suggests that language-use conventions vary across ethnic
groups in the same country, for instance in the United States (cf., e.g. Wolfram &
Schilling-Estes 2006:93102). So far, however, systematic empirical studies of ethnic
differences are extremely rare. Clancy (forthcoming) is an exception. Clancy analyzes
the use of hedges in family conversation from an Irish middle-class family and a family
from the Irish traveller community, both in the Limerick area (Irish travellers are an
ethnic community comparable to the Sinti and Roma on the European continent).
That no work on socioeconomic variation exists in variational pragmatics to date
is not surprising considering the fact that of all the macro-social factors investigated,
social class is the most undertheorized factor (cf. Section 3.1 above). Some studies
do, however, refer to social-class membership, yet there is no direct comparison. For
example, Macaulay (2009) does study working-class adolescents, but he does not contrast them with middle-class adolescents; his focus is on gender and age variation.
Of the five levels of pragmatic analysis distinguished in variational pragmatics
(cf. Section 2.3 above), it is the actional level which has received more attention
than any other level. This is not surprising, since speech act-based studies generally
dominate in most branches of empirical pragmatics (cf. Section2.2 above). The speech
acts which have been examined in variational pragmatics include, first and foremost,
requests (as in other branches of empirical pragmatics) (cf., e.g. Barron 2008; Muhr
2008; Warga 2008; Placencia 2008; Flix-Brasdefer 2009). Further speech acts include
apologies (e.g. Schlmberger 2008), offers (e.g. Barron 2005a), invitations (e.g. Garca
2008), compliment responses (e.g. Schneider 1999), thanks (e.g. Jautz 2008), responses
to thanks (e.g. Schneider 2005a), and reprimands (e.g. Garca 2009). In most cases,
pragmalinguistic aspects are compared, i.e. aspects of speech act realization: either the
conventions of form (i.e. lexico-grammatical choices; cf., e.g. Schneider 2005a; Jautz
2008), or the conventions of means (i.e. the strategy types selected; cf., e.g. Warga 2008;
Schlmberger 2008), or modification patterns (cf., e.g. Barron 2008). As a rule, for this
type of study on the actional level of analysis the framework developed in the CCSARP
(Blum-Kulka et al. 1989b) is adopted for analysing the data, and Brown & Levinsons
politeness theory (1987) for interpreting the results. Sociopragmatic parameters have
been researched to a lesser extent.
The second area which has been investigated relatively intensively is the area
of discourse markers (in a broad sense of the term). Examples include Id say, I say,
I mean and you know (Kallen 2005), I dont know and I dont think (Pichler 2009),

Variational pragmatics 259

hedges such as like, actually and I think (Clancy forthcoming), response tokens such as
yeah, exactly, is that right? (OKeeffe & Adolphs 2008), and religious references such as
God Almighty and Jesus Christ (Farr & Murphy 2009). As a rule, discourse markers are
examined on the formal level, i.e. the analysis starts with a form and is aimed at determining the different communicative and social functions of this form (cf. Section2.3
above). In several cases, however, the opposite function-to-form mapping approach
is adopted, which is characteristic of the analyses on the actional level, but used on
the formal level with discourse units smaller than speech acts, i.e. individual words or
phrases. An example would be the analysis of hedging which is aimed at establishing
which expressions can be used in this function.
Work on the remaining three levels of analysis currently distinguished in variational
pragmatics the interactional, the topic and the organisational level (cf. Section2.3
above) is extremely rare. Examples are Schneiders studies of small talk in English
English, Irish English and American English (Schneider 2008a and in press), in which
complete dialogues are analysed. Aspects which are examined include openings and
closings, illocutions and move structure, sequencing and topic selection.
The methods employed in variational pragmatics to date are, first and foremost, production questionnaires and corpora. While questionnaires are used predominantly for
studies of the actional level, corpora are overwhelmingly used for studies of the formal
level. This is not surprising, since pragmatically annotated corpora are not yet available (cf. Section4.2 above) and corpus searches are considerably easier for individual
words or invariant phrases than for speech act realization patterns (cf., e.g. Jucker
2009). In most cases, the questionnaires used are written questionnaires which comprise discourse completion tasks, in which the informants are requested to provide
only one turn (cf., e.g. Barron 2008; Warga 2008; Schlmberger 2008). In some cases,
however, dialogue production tasks are used, in which the informants have to provide
a complete dialogue (cf., e.g. Barron 2005a; Schneider 2008a). The corpora on which
studies have been based include the well-known British National Corpus (BNC) and
the Wellington Corpus of Spoken New Zealand English. Interestingly, in Jautzs work
on thanking, which is based on these two corpora (cf., e.g.Jautz 2008 and 2010), the
approach primarily adopted is form-to-function mapping, i.e. the approach usually
adopted in studies of the formal level (cf. Section2.3 above). The corpora best suited
for the purposes of variational pragmatics are corpora which are structured according
to the same principles. Examples are the components of the International Corpus of
English (ICE), for instance the British and the Irish components, ICE-GB and ICEIreland, analysed, e.g. by Kallen (2005). Further examples are the Cambridge and
Nottingham Corpus of Discourse in English (CANCODE) and the Limerick Corpus
of Irish English (LCIE), which is modelled on CANCODE. A study employing these
two corpora is OKeeffe & Adolphs (2008); LCIE is also used in OKeeffe & Amador
Moreno (2009), Farr & Murphy (2009) and Murphy (2010). In their comparison of

260 Klaus P. Schneider

Dutch spoken in the Netherlands and in Flanders, Plevoets et al. (2008) extract their
data from a corpus of spoken Dutch called Corpus Gesproken Nederlands (CGN),
which comprises a Netherlandic as well as a Belgian component. The data source
analysed in Placencia (2008) is also a corpus of naturally occurring discourse. In this
case, however, the audio-recordings were made specifically for the purpose of that
particular analysis. Recently, a case has been made for the compilation and use of
small self-recorded up-to-date multi-modal corpora, for which not only transcripts,
but also audio-visual data and comprehensive demographic information are available
(cf. Clancy forthcoming). It has been pointed out that such corpora, which can be
focused on one particular genre or one community of practice, is best suited to meet
the methodological requirements of variational pragmatics (cf. Section4.1 above).
Further methods of data collection employed in variational pragmatics include
interviews and role play. So far, role plays have been made use of especially in speechact based studies across national varieties of Spanish (cf., e.g. Garca 2008 and 2009;
Flix-Brasdefer 2009), whereas interviews have been used especially in studies
focused on the interaction of social factors in one community (cf., e.g. Macaulay
2009; Pichler 2009). The use of role plays rather than production questionnaires in
studies of the actional level permits the inclusion of prosody in the analysis. Unfortunately, prosodic features and intonation have been largely neglected to date (cf., however, Flix-Brasdefer 2009).
Finally, it is worth noting that the vast majority of studies mentioned in the
present section are studies explicitly placed in the framework of variational pragmatics.
Needless to say, there may be other studies outside this explicit framework which are
related in that they compare pragmatic phenomena across varieties of the same language (as described in Section2.2 above).
6. Perspectives
Ideally, all levels of analysis and all macro-social factors specified in the present framework of variational pragmatics (cf. Sections2.3 and 3.1 above) should receive the same
degree of attention in research. However, this is not yet the case, as has been shown in
Section5.2. Therefore, it is necessary to carry out more empirical projects on pragmatic
phenomena more complex than discourse markers and speech acts as well as on ethnic
and socioeconomic variation. Furthermore, it may be necessary to include additional
macro-social factors which may be found to systematically impact verbal behaviour
and may be identified in studies of languages so far neglected. Candidates of such
factors might be education and religion, and languages so far neglected include, by
and large, all non-Indo-European languages (which is not to say that Indo-European
languages have been covered sufficiently). Moreover, more effort should be invested

Variational pragmatics

in examining the interplay of many and, eventually, all social factors, including microsocial factors. Finally, it seems desirable to integrate into the present framework
relevant concepts from other traditions, e.g. register, activity type and community of
practice (cf. Section3.2 above). Two concepts of this kind are the focus of Schneider
(2007) and Haugh & Schneider (forthcoming).
Up to now, pragmatic phenomena have been examined more or less in isolation.
For instance, a number of individual speech acts have been analysed, but no attempts
have been made to synthesize the findings on different speech acts investigated in the
same variety, e.g. a national variety of a pluricentric language. The overarching question in this case is whether there are any general features which are variety-specific
and therefore can be observed in the realization of different speech acts in that same
variety. Questions of this type have been addressed only in passing. For instance, it was
found that across pragmatic phenomena on various levels of analysis, i.e. including
speech acts but not restricted to speech acts, a high degree of indirectness is characteristic to Irish English by comparison to other national varieties of English, notably
English English (cf. Barron & Schneider 2005a: 5, Kallen 2005).
A further concern is to go beyond the identification and description of pragmatic
differences across varieties of the same language and to address the question of why
the observed differences occur. In other words, an explanation for the existence of
pragmatic variation is sought. It has been claimed that some of the differences identified (for instance, opening a conversation with a stranger at a social gathering) can
be attributed to underlying schemata which are culturally determined (cf.Schneider
2010). These schemata are cognitive structures in the long-term memory which can be
referred to as cultural models or cultural scripts (but not necessarily as cultural scripts
in the ethnopragmatic sense; cf., e.g. Goddard & Wierzbicka 2007 and Section2.2
above). Also, attempts have been made to account for cross-varietal variation by
referring to models for the explanation of general differences between cultures, or more
specifically national cultures, developed in business studies or the social sciences.
Examples are Hofstedes well-known model (cf., e.g. Hofstede 1980) and also the
more recent and more refined model developed in the international GLOBE project
(cf.,e.g.House etal. 2002). However, these projects have been found to be of only
limited value (cf.,e.g.Barron 2008).
It is worth noting that research in variational pragmatics has overwhelmingly,
if not exclusively, dealt with spoken discourse. There is, however, reason to believe
that pragmatic variation across varieties of the same language also exists in written
discourse, including computer-mediated communication (e.g. e-mails and blogs).
In all likelihood, this type of pragmatic variation does not occur in all written genres,
but at least some genres can be expected to be susceptible to cultural influences.
It can be assumed that for research in this new area the framework presented in
the present article (cf. especially Section2.3 above) has to be adapted and modified

261

262 Klaus P. Schneider

based on insights from text linguistics, genre analysis and discourse analysis (possibly
including critical discourse analysis) (cf., e.g. Brinker 2010; Bhatia 2004; Johnstone
2008; Fairclough 2003).
Further perspectives for variational pragmatics pertain to its social relevance and
significance for society (cf. Schneider & Barron 2008a: 2122). Diverging expectations
and behaviours in verbal interaction may result in serious misunderstandings, social
tensions and communication breakdown. Wolfram & Schilling-Estes (2006:101), writing about pragmatic variation in the United States, emphasize that some of the major
areas of social dissonance and conflict among social and ethnic groups in American
society are directly tied to peoples failure to understand that different groups have
different language-use conventions. Therefore it seems necessary to introduce findings from variational pragmatics in language teaching and communication trainings,
including foreign language teaching and intercultural trainings, to develop the pragmatic competence of language users and their awareness of diverging language-use
conventions across regional and social varieties of the same language (cf. Schneider
2008b). Teaching materials and textbooks for these types of training have yet to be
written. In her review of Schneider & Barron (2008b), Tottie (2009) underscores this
practical applicability of variational pragmatics. She concludes (Tottie 2009: 525):
When results from this research become generally known, for example via textbooks,
we can hope for fewer pragmatic failures in communication. It is not only for this
reason, however, that it is hoped that Tottie is right when she says (2009:525): I am
convinced that variational pragmatics has a bright future.

References
[A continuously updated bibliography of variational pragmatics is found at: www.linguistics.uni-bonn.
de/research/variational-pragmatics/variational-pragmatics-bibliography].
Ash, S. (2002). Social class. In J.K. Chambers, P. Trudgill & N. Schilling-Estes (eds.) Handbook of
language variation and change (Blackwell Handbook in Linguistics): 402422. Blackwell.
Barron, A. (2005a). Offering in Ireland and England. In A. Barron & K.P. Schneider (eds.) (2005b):
141176.
Barron, A. (2005b). Variational pragmatics in the foreign language classroom. System 33(3): 519536.
Barron, A. (2008). The structure of requests in Irish English and English English. In K.P. Schneider&
A. Barron (eds.) (2008b): 3567.
Barron, A. (2009). Apologies across the U.S.A. In K. Turner & B. Fraser. (eds.) Language in life, and a
life in language. Jacob Mey A Festschrift: 917. Emerald.
Barron, A. & K.P. Schneider (2005a). Irish English: a focus on language in action. In A. Barron &
K.P. Schneider (eds.) (2005b): 315.
Barron A. & K.P. Schneider (eds.) (2005b). The pragmatics of Irish English. (Trends in Linguistics.
Studies and Monographs 164). Mouton de Gruyter.
Barron, A & K.P. Schneider (2009a). Variational pragmatics: studying the impact of social factors on
language use in interaction. Intercultural Pragmatics 6(4): 425442.

Variational pragmatics

Barron, A. & K.P. Schneider (eds.) (2009b). Variational pragmatics. Special Issue. Intercultural
Pragmatics 6(4).
Bhatia, V.K. (2004). Worlds of written discourse: a genre-based view. Continuum.
Blum-Kulka, S., J. House & G. Kasper (1989a). Investigating cross-cultural pragmatics: an introductory
overview. In S. Blum-Kulka, J. House & G. Kasper (eds.) Cross-cultural pragmatics: requests and
apologies (Advances in Discourse Processes 31): 134. Ablex.
Blum-Kulka, S., J. House & G. Kasper (eds.) (1989b). Crosscultural pragmatics: requests and
apologies. Ablex.
Blum-Kulka, S., M. Blondheim, J. House, G. Kasper & J. Wagner (2008). Intercultural pragmatics,
language and society. In P. van Sterkenburg (ed.): Unity and diversity of languages: 155173.
Benjamins.
Brinker, K. (2010). Linguistische Textanalyse. 7th revised edition. Erich Schmidt.
Brown, G. & G. Yule (1983). Discourse analysis. Cambridge University Press.
Brown, P. & S.C. Levinson (1978). Universals in language usage: politeness phenomena. In
E.N.Goody (ed.) Questions and politeness: strategies in social interaction: 56289. Cambridge
University Press.
Brown, P. & S.C. Levinson (1987). Politeness: some universals in language usage. Cambridge
University Press.
Cheshire, J. (2002). Sex and gender in variationist research. In J.K. Chambers, P. Trudgill &
N.Schilling-Estes (eds.) Handbook of language variation and change (Blackwell Handbook in
Linguistics): 423443. Blackwell.
Clancy, B. (forthcoming). I now pronounce you man and wife: exploring the synergy of variational
pragmatics and corpus linguistics. In F. Farr & A. OKeeffe (eds.) Applying corpus linguistics.
Special issue. International Journal of Corpus Linguistics.
Clark, H.H. (1979). Responding to indirect speech acts. Cognitive Psychology 11: 430477.
Clyne, M. (1992). Introduction. In M. Clyne (ed.) (1992) Pluricentric languages: Different norms in
different countries: 19. Mouton de Gruyter.
Downes, W. (1998). Language and society. Cambridge University Press.
Edmondson, W.J. (1981). Spoken discourse: a model for analysis. Longman.
Ervin-Tripp, S. (1976). Is Sybil there? The structure of some American English Directives. Language
in Society 5: 2566.
Faerch, C. & G. Kasper (eds.) (1987). Introspection in second language research. Multilingual Matters.
Fairclough, N. (2003). Analysing discourse: Textual analysis for social research. Routledge.
Farr, F. & B. Murphy (2009). Religious references in contemporary Irish English: for the love of God
almighty. Im a holy terror for turf. In A. Barron & K.P. Schneider (eds.) (2009b) Variational
Pragmatics. Special Issue. Intercultural Pragmatics 6(4): 535559.
Flix-Brasdefer, J.C. (2009). Pragmatic variation across Spanish(es): requesting in Mexican, Costa
Rican and Dominican Spanish. Intercultural Pragmatics 6(4): 473515.
Garca, C. (2008). Different realizations of solidarity politeness: comparing Venezuelan and Argentinean
invitations. In K.P. Schneider & A. Barron (eds.) (2008b): 269305.
Garca, C. (2009). Intra-lingual pragmatic variation in the performance of reprimanding. Intercultural
Pragmatics 6(4): 443472.
Goddard, C. (2006). Ethnopragmatics. A new paradigm. In C. Goddard (ed.) Ethnopragmatics:
understanding discourse in cultural context (Applications in Cognitive Linguistics 3): 130.
Mouton de Gruyter.
Goddard, C. (2009). Cultural keywords and cultural scripts: divergences between Australian, American
and British English. Presentation at the 11th International Pragmatics Conference, University of
Melbourne, Australia, 1217 July 2009.

263

264 Klaus P. Schneider

Goddard, C. & A. Wierzbicka (2007). Semantic primes and cultural scripts in language learning
and intercultural communication. In F. Sharifian & G. Palmer (eds.) Applied cultural linguistics:
105124. Benjamins.
Golato, A. (2002). German compliment responses. Journal of Pragmatics 34: 547671.
Golato, A. (2003). Studying compliment responses: a comparison of DCTs and recordings of naturally
occurring talk. Applied Linguistics 24: 90121.
Goodwin, M.H. (1990). He-said-she-said: Talk as social organization among black children. Indiana
University Press.
Haugh, M & K.P. Schneider (eds.) (in preparation). (Im)politeness across Englishes. Special Issue. Journal
of Pragmatics.
Hill, B., Ide, S., Ikuta, S., Kawasaki, A. & Ogino, T. (1986). Universals of linguistic politeness. Journal
of Pragmatics 10: 347371.
Hofstede, G.H. (1980). Cultures consequences: international differences in work-related values. Sage.
Holmes, J. (1986). Compliments and compliment responses in New Zealand English. Anthropological
Linguistics 28: 485508.
Holmes, J. (1990). Apologies in New Zealand English. Language in Society 19(2): 155199.
Holmes, J. (1995). Women, men and politeness. Longman.
House, J. (2003). English as a lingua franca and its influence on discourse norms in other languages.
In G. Anderman & M. Rogers (eds.) Translation today: trends and perspectives: 168179.
Multilingual Matters.
House, R.J., M. Javidan, P. Hanges & P. Dorfman (2002). Understanding cultures and implicit leadership
theories across the globe: an introduction to project GLOBE. Journal of World Business 37: 310.
Huang, Y. (2007). Pragmatics. Oxford University Press.
Huang, Y. (2010). Pragmatics. In L. Cummings (ed.) The pragmatics encyclopedia: 341345. Routledge.
Janney, R. (2009). Toward a postcolonial pragmatics. In K. Turner & B. Fraser (eds.) Language in life,
and a life in language. Jacob Mey A Festschrift: 193202. Emerald.
Jautz, S. (2008). Gratitude in British and New Zealand radio programmes: Nothing but gushing?
In K.P. Schneider & A. Barron (eds.) (2008b): 141178.
Jautz, S. (2010). Thanking in English: a corpus-based study on the use of expressions of gratitude across
different varieties and genres. Unpublished Ph.D. dissertation. University of Bonn.
Johnstone, B. (2008). Discourse analysis. 2nd edition. Blackwell.
Jucker, A.H. (ed.) (1995). Historical pragmatics: pragmatic developments in the history of English.
Benjamins.
Jucker, A.H. (2008). Historical pragmatics. Language and Linguistics Compass 2(5): 894906.
Jucker, A.H. (2009). Speech act research between armchair, field and laboratory: the case of compliments. Journal of Pragmatics 41: 16111635.
Jucker, A.H. & I. Taavitsainen (eds.) (2008). Speech acts in the history of English. Benjamins.
Jucker, A.H. & I. Taavitsainen (forthcoming). Pragmatic variables. In J.M. Hernndez-Campoy (ed.)
The handbook of historical sociolinguistics. Blackwell.
Kachru, B. (1985). Standards, codification and sociolinguistic realism: the English language in the
outer circle. In R. Quirk & H.G. Widdowson (eds.) English in the world: teaching and learning of
language and literature: 1136. Cambridge University Press.
Kallen, J.L. (2005). Silence and mitigation in Irish English discourse. In A. Barron & K.P. Schneider
(eds.) The pragmatics of Irish English. (Trends in Linguistics. Studies and Monographs 164):4771.
Mouton de Gruyter.
Kasper, G. (1995). Wessen Pragmatik? Fr eine Neubestimmung fremdsprachlicher Handlungskompetenz. Zeitschrift fr Fremdsprachenforschung 6(1): 6994.

Variational pragmatics 265

Kasper, G. (2000). Data collection in pragmatics research. In H. Spencer-Oatey (ed.) Culturally speaking: Managing rapport through talk across cultures: 316341. Continuum.
Kasper, G. (2008). Data collection in pragmatics research. In H. Spencer-Oatey (ed.) Culturally speaking:
managing rapport through talk across cultures. 2nd edition: 279303. Continuum.
Labov, W. (1966a). The linguistic variable as a structural unit. Washington Linguistics Review 3: 422.
Labov, W. (1966b). The social stratification of English in New York City. Center for Applied Linguistics.
Lavandera, B. (1978). Where does the sociolinguistic variable stop? Language in Society 7(2):
171182.
Leech, G. (1983). Principles of pragmatics. Longman.
Levinson, S.C. (2000). Presumptive meanings: the theory of generalized conversational implicature.
MIT Press.
Locher, M. (2008). Relational work, politeness and identity construction. In: G. Antos,E. Ventola&
T. Weber (eds.) Handbooks of applied linguistics: interpersonal communication: 509540.
Mouton de Gruyter.
Locher, M. & R. Watts (2005). Politeness theory and relational work. Journal of Politeness Research
1: 933.
Macaulay, R. (2009). Adolescents and identity. In A. Barron & K.P. Schneider (eds.) Variational
pragmatics. Special Issue. Intercultural Pragmatics 6(4): 597612.
Mrquez Reiter, R. (2002). A contrastive study of conventional indirectness in Spanish: evidence
from Peninsular and Uruguayan Spanish. Pragmatics 12(2): 135151.
McLaughlin, M.L. (1984). Conversation: how talk is organized. Sage.
Melchers, G. & P. Shaw (2003). World Englishes: an introduction. Arnold.
Milroy, J. (1981). Regional accents of English: Belfast. Blackstaff Press.
Milroy, J. (1998). Urban dialectology. In J. Mey (ed.) Concise encyclopedia of pragmatics: 10391045.
Elsevier.
Milroy, L. (1980). Language and social networks. Blackwell.
Muhr, R. (1994). Entschuldigen Sie Frau Kollegin: Sprechaktrealisierungsunterschiede an
Universitten in sterreich und Deutschland. In G. Bachleitner-Held (ed.) Verbale Interaktion:
126143. Dr. Kovac.
Muhr, R. (2008). The pragmatics of a pluricentric language: a comparison between Austrian
German and German German. In K.P. Schneider & A. Barron (eds.) (2008b): 209242.
Murphy, B. (2010). Corpus and sociolinguistics: investigating age and gender in female talk. Benjamins.
OKeeffe, A. & S. Adolphs (2008). Response tokens in British and Irish discourse: corpus, context
and variational pragmatics. In K.P. Schneider & A. Barron (eds.) (2008b): 6998.
OKeeffe, A. & C.P. Amador Moreno (2009). The pragmatics of the be + after + V-ing construction
in Irish English. Intercultural Pragmatics 6(4): 517534.
Oleksy, W. (ed.) (1989). Contrastive pragmatics. Benjamins.
Pichler, H. (2009). The functional and social reality of discourse variants in a nothern English dialect:
I DONT KNOW and I DONT THINK compared. Intercultural Pragmatics 6(4): 561596.
Placencia, M.E. (1994). Pragmatics across varieties of Spanish. Donaire 2: 6577.
Placencia, M.E. (2008). Requests in corner shop transactions in Ecuadorian Andean and Coastal
Spanish. In A. Barron & K.P. Schneider (eds.) (2008b): 307332.
Plevoets, K., D. Speelman & D. Geeraerts (2008). The distribution of T/V pronouns in Netherlandic and
Belgian Dutch. In A. Barron & K.P. Schneider (eds.) (2008b): 181209.
Sacks, H., E.A. Schegloff & G. Jefferson (1974). A simplest systematics for the organization of turntaking for conversation. Language 50: 696735.
Schiffrin, D. (1987). Discourse markers. Cambridge University Press.

266 Klaus P. Schneider

Schlieben-Lange, B. & H. Weydt (1978). Fr eine Pragmatisierung der Dialektologie. Zeitschrift fr
Germanistische Linguistik 6(3): 257282 [pp. 257260 are given twice in this volume, each time
with different content].
Schneider, K.P. (1988). Small talk: analysing phatic discourse. Hitzeroth.
Schneider, K.P. (1999). Compliment responses across cultures. In M. Wysocka (ed.) On language
in theory and practice: In honour of Janusz Arabski on the occasion of his 60th birthday Vol. 1:
162172. Wydawnictwo Uniwersytetu lskiego.
Schneider, K.P. (2001). Pragmatics meets dialectology: investigating regional variation in language
use. Invited paper presented at the international workshop New Orientations in the Study of
Regionalism, University of Bonn, Germany, July 16th19th 2001.
Schneider, K.P. (2005a). No problem, youre welcome, anytime: responding to thanks in Ireland,
England, and the U.S.A. In A. Barron & K.P. Schneider (eds.): The pragmatics of Irish English:
101139. Mouton de Gruyter.
Schneider, K.P. (2005b). Region and regionalism in linguistics: a brief survey of concepts and methods.
In L. Hnnighausen, M. Frey, J. Peacock & N. Steiner (eds.) Regionalism in the age of globalism:
concepts of regionalism 1: 139158. University of Wisconsin, Center for the Study of Upper
Midwestern Cultures.
Schneider, K.P. (2007). Genre matters: textual and contextual constraints on contemporary English
speech behaviour. Anglia 125 (1) Special Issue: Spoken styles of present-day English): 5983.
Schneider, K.P. (2008a). Small talk in England, Ireland, and the USA. In K.P. Schneider & A. Barron
(eds.) (2008b): 99139.
Schneider, K.P. (2008b). Teaching intercultural competence and pragmatic variation. In E. BurwitzMelzer et al. (eds.) Sprachen lernen Menschen bilden: 179189. Schneider-Verlag Hohengehren.
Schneider, K.P. (2010). Sociopragmatic variation and culture-dependent schemata of linguistic
behavior. LAUD paper A748. University of Duisburg-Essen, Linguistic Agency.
Schneider, K.P. (in press). Small talk: units, sequencing, realizations. In A. Wimmer & J. Helbig (eds.)
Anglistentag 2009 Klagenfurt: Proceedings. WVT.
Schneider, K.P. (forthcoming). Imagining conversation: how people think people do things
with words. Sociolinguistic Studies. Special Issue: Fictionalising orality (eds. K. McCafferty,
C.Amador-Moreno).
Schneider, K.P. & A. Barron (2008a). Where pragmatics and dialectology meet: introducing variational
pragmatics. In K.P. Schneider & A. Barron (eds.) (2008b): 132.
Schneider, K.P. & A. Barron (eds.) (2008b). Variational pragmatics: a focus on regional varieties in
pluricentric languages (Pragmatics & Beyond New Series 178). Benjamins.
Schlmberger, U. (2008). Apologizing in French French and Canadian French. In K.P. Schneider &
A. Barron (eds.) (2008b): 329350.
Selinker, L. (1972). Interlanguage. IRAL 10: 209231.
Sridhar, K.K. (1991). Speech acts in an indigenised variety: sociocultural values and language variation.
In J. Cheshire (ed.) English around the world: sociolinguistic perspectives: 308318. Cambridge
University Press.
Terkourafi, M. (forthcoming). The pragmatic variable: a procedural approach. Language in Society.
Thomas, J. (1981). Pragmatic failure. Unpublished M.A. thesis. University of Lancaster.
Tottie, G. (2009). Review of Schneider & Barron 2008: Variational pragmatics. Language in Society
38(4): 520525.
Trosborg, A. (1995). Interlanguage pragmatics. Mouton de Gruyter.
Trudgill, P. & J. Hannah (2002). International English: a guide to the varieties of Standard English
(4thed.). Arnold.

Variational pragmatics

Verschueren, J. (1999). Understanding pragmatics. Arnold.

Warga, M. (2008). Requesting in German as a pluricentric language. In K.P. Schneider & A. Barron
(eds.) (2008b): 243264.
Watts, R. (2003). Politeness. Cambridge University Press.
Wierzbicka, A. (1985). Different cultures, different languages, different speech acts: Polish vs. English.
Journal of Pragmatics 9(2/3): 145178.
Wierzbicka, A. (1991). Cross-cultural pragmatics: the semantics of human interaction. Mouton
de Gruyter.
Wolfram, W. & N. Schilling-Estes (1998). American English: dialects and variation. Blackwell.
Wolfram, W. & N. Schilling-Estes (2006). American English: dialects and variation (2nd ed.). Blackwell.
Wolfson, N. (1988). The bulge: a theory of speech behaviour and social distance. In J. Fine (ed.)
Second language discourse: A textbook of current research: 2138. Ablex.

267

Index

A
Aarts, Bas, 180, 185
Aboh, E.O., 52, 54
Acheulean industry, 91
acrolect, 55
Adam, L., 51
Adolphs, S., 259
Adorno, T., 167
African-American English
(AAE), 51, 66
African-American vernacular
English (AAVE), 49, 66
Ager, D., 224
Aijmer, K., 131
Aikhenvald, A.Y., 2627
Aitchison, J., 171
Albert, M.L., 204
Alleyne, M.C., 48, 51
Allsopp, R., 52
Allwood, J., 187
Amador Moreno, C.P., 256, 259
American English, 66
American Sign Language
(ASL), 23
Andersen, H., 4, 106, 185
Andersen, R., 55
Andreini, L., 167
Androutsopoulos, J., 38
Ansaldo, U., 54
antilanguage, 160162
Anttila, R., 101, 108
Appalachian English, 67
appeal, 83
Appel, R., 19, 2125
approaches to variation and
change, 25
areal diffusion, 211
areal linguistics, 38
Arends, J., 5455
Arndt, H., 150
Arnovick, L.K., 116
Arriv, M., 225
Arveiller, R., 48
Ash, S., 70, 248

asymmetrical
communication, 249250
Atlas, J.D., 124126, 129, 139
Atlas linguistique de France, 79
Auer, P., 28, 41, 206
Av-Lallemant, F.C.B., 157
B
Backus, A., 22
Baetens Beardsmore, H., 24
Bailey, C.J., 63
Baissac, C., 51
Baker, P., 4849, 54
Bakker, P., 27, 41, 211212
Baldi, P., 103, 172
Barcelona, A., 186
Bardovi-Harlig, K., 152
Barron, A., 40, 239, 241, 244,
246,251252, 254259,
261262
basilect, 55
Bauer, L., 171
Beard, A., 171
Becker, P.R., 89
Beier, L., 156
Berger, P., 164
Bergs, A., 7
Bernstein, Basil, 81
Berruto, G., 231
Bertrand-Bocand, E., 51
Bertuccelli Papi, M., 115
Bhatia, V.K., 262
Bialystok, E., 145146
Biber, D., 192
Bickerton, D., 4647, 49, 53, 63,
106, 210
Bilingual First Language
Acquisition, 204
bilingual individual, 204205
BIMOLA (Bilingual Mode of
Lexical Access), 31
Bisang, W., 67
Black English Vernacular
(BEV), 63

Blake, N.F., 114

Blanche-Benveniste, C.,
224 225, 231
Blommaert, J., 28, 206
Bloomfield, L., 18, 21, 4849,
61, 173
Blum-Kulka, S., 142, 153, 242243,
245, 250, 254, 258
Boberg, C., 70
Boesch, C., 87, 89
Bolonyai, A., 207
Bolton, K., 40
borrowing, classifications
of, 2123
borrowing, constraints on, 2529
borrowing, constraints problem
on, 17
borrowing, definitions of, 1921
borrowing, historical
overview, 1819
borrowing and
code-switching, 21
borrowing-proof, 26
borrowing scale, 26
Bou Franch, P., 40
Bouton, L.F., 152
Breuer, H., 117
Brinker, K., 262
Brinton, L., 171
Brinton, L.J., 78, 110, 114
Brisard, F., 42, 81
Britain, D., 191
British and the Irish components
of the International Corpus
of English, 253
British National Corpus, 253
Brophy, J., 166
Brown, G., 246
Brown, P., 52, 111, 116, 237, 241,
250, 258
Brown, R., 116
Brugmann, K., 215
Bhler, K., 8384
Burke, P., 156, 167

Index 269

Busse, U., 12, 118
Bybee, J., 6, 131133, 138
Bybee, J.L., 106
Bynon, R.T., 171
Bynon, T., 171
C
Cadiot, P., 62
Cambridge and
Nottingham Corpus of
Discourse in English
(CANCODE), 253,259
Cameron, D., 62
Campbell, L., 42, 105, 171, 173,
182, 187188, 190,
197198,208
Canadian English, 66
Carter, H., 52
Cassidy, F.G., 69
Cassirer, E., 83, 87
Cavendar, A., 65
Chafe, W., 3
Chambers, J.K., 6163, 67,171
Chaudenson, R., 4648, 5354
Chen, M., 186
Cheshire, J., 183, 248
Chinook Jargon, 47
Chirsheva, G., 24
Chomsky, N., 8182, 172
Chomskyan linguistics, 103
Choron-Baix, C., 166
Clancy, B., 253, 258260
Clark, H.H., 245
Clark, R., 184
Clyne, M., 20, 39, 205, 248
code alternation, 205
code copying, 20
code-mixing, 205
codes, 228
code-switching, 19, 205
Coelho, F.A., 53
cognitive ecology, 93
Cohen, A., 143
Colantoni, L., 19
Collins, D.E., 113
communicational
dialectology, 75
communicative effect, 149150
communicative style, 225
complementary hypothesis, 54
Composite Matrix
Language, 207
compound bilinguals, 204
Comrie, B., 211, 234237

conative function, 83
congruent lexicalization, 205
contact accounts of language
change, 194197
contact and location, 3738
contact as process, 4243
contact-induced language
change, 207210
contact linguistics, 36, 38
contact linguistics, direction of
interference, 3841
contact linguistics, indirect
influence in language
contact, 42
contextualization cues, 206,
225226
contrastive pragmatics, 242
conventional implicature,
123124
conventions of form, 245
conventions of means, 245
conversational implicature, 124
conversational principles,
124131
Cooperative Principle, 124
coordinate bilinguals, 204
Corne, C., 49, 54
Corpus Gesproken Nederlands
(CGN), 260
Corpus of Early English
Correspondence
(CEEC), 112
Corpus of English Dialogues, 113
Coulmas, F., 144
Coupland, N., 42, 150
Coyaud, M., 167
creole and creolization, 50
creole genesis, 52
creoles, 46, 48
creolized language, 48
Crisma, P., 182
Croft, W., 2, 5, 174, 234
Cro-Magnon life space and
pragmatic space, 9395
cross-cultural pragmatics, 242, 244
Cross-Cultural Speech Act
Realization Project
(CCSARP), 254
cross-language variation, 1013
cross-modal borrowing, 23
Crowley, T., 171
Crum, M., 52
Crystal, D., 40, 204
Culbertson, R.G., 156

Culpeper, J., 112113, 117

cultural borrowing, 21
cultural contact, 195
cultural scripts, 243
D
Dahl, M., 38, 106, 152
Dahl, ., 38, 106, 152
Darwin, C., 81
Dasher, R., 171
Dasher, R.B., 2, 6, 171, 187
Dauzat, A., 157, 166
Davidson, I., 88, 9193
Dawkins, R.M., 24, 27, 209
DCosta, J., 49
de Chaves, C., 157
Delaware Pidgin, 47
Deleuze, G., 159160, 162
delinquent jargon, differentiation
and opposition, 162163
delinquent jargon, integration
and creation of private
world, 163164
delinquent jargon,
liberation, 164165
delinquent jargon,
stigmatizing, 165
delinquent jargon and
value-giving measure,
156160
DEloia, S.G., 53
Denison, D., 185, 191
De Rooij, V.A., 19, 28
Descamps-Hocquet, M., 168
Deuchar, M., 41
Deulofeu, J., 224225, 231
Deutschmann, M., 115
diachronic analysis and
pathways of change, 58
diagrammatic iconicity, 175
dialect, 61
dialect atlases, 79
dialect borrowing, 21
dialect continuum, 62
dialect differences, 61
dialect maps, 79
dialect of Ayr, 67
dialectology, communicative
aspects, 7879
dialectology, definition, 73
dialectology, history, 7375
dialectology, impact of history
on origin and evolution of
languages, 76

270 Index
dialectology, linguistic
reconstruction, 7677
dialectology, sociopragmatic and
attitudinal aspects, 7778
dialectology, study of universals
of language, 77
dialectology in U.S., recent
developments to, 6870
dialects as local forms of
speech, 6467
Dialects in Contact, 65
dialects of Anniston, 67
dialects of Glasgow, 67
dialect studies, examples
of, 6667
dialect vs. register, 226227
diasystem, 73
Dictionary of American Regional
English (DARE), 6870
Diewald G., 7, 179
Dik, S.C., 106
Dillard, J.L., 53
Dirckx, J.H., 166
direct/imagic iconicity, 174
directive and assertive
meanings, 128129
Dirven, R., 187
discourse markers, 113114
discourse mode, 222223
discourse-related
code-switching, 206
Dittmar, N., 221232
Dixon, R.M.W., 27
Dixon, R.W., 105
Dorian, N., 40, 42, 196, 204
Dorian, N.C., 173
Dorleijn, M., 22
Dossena, M., 112
Downes, W., 248
Dryer, M., 236
Dryer, M.S., 236
Dunbar, R., 87
Durie, M., 188
E
E.C. Traugott, 23, 67,
106107, 110, 126129, 133135,
137, 171, 187188, 192193
eclectic, 244
Edmondson, W.J., 144, 245
Edwards, V., 67
Ehrismann, G., 111, 117
embedding problem,
borrowing, 17

Emeneau, M.B., 38
emergence of sign
functions, 86
emerging metaphors, 137
emotive function, 83
epistemic meanings,
development of, 128129
Equivalence Constraint, 206
Ervin, S., 204
Ervin-Tripp, S., 241
ethnopragmatics, 243
etymon, 216
evolution of tool-use, 8891
expanded pidgins, 46
expression, 83
F
Faerch, C., 252
Faine, J., 53
Fairclough, N., 262
false friends, 41
Farr, F., 258259
Feagin, C., 67
Fehr, E., 86
Flix-Brasdefer, J.C.,
256258, 260
Fennell, B., 171
Ferguson, C., 223, 229
field (discursive speech-usage
field), 222
Filppula, M., 195
Finkenstaedt, Th., 111, 117
Firth, J.R., 221
Fischer, O., 6, 174, 179, 182184,
188, 193, 198
Fishman, J., 40, 203
Fitch, W.T., 4, 81
Fitzmaurice, S.M., 111112
folk dialectology, 68
folk linguistics, 68
Fortune, E.P., 156
Foucault, M., 156157, 159, 164
Fox, A., 194
Francis, W.N., 61
Franois, D., 155156, 159
Franois-Geiger, D., 168
Freeman, D., 40
Freeman, Y., 40
Free Morpheme Constraint, 206
Fried, M., 114
Fritz, G., 111
fudged lects, 63
functional linguistics la
Ferguson, 229230

G
Gadet, F., 13, 225
Gal, S., 204
Galmiche, M., 225
Garca, C., 256258, 260
Garca Ramos, J., 161
Grdenfors, P., 187
Gardner-Chloros, P., 205
Gaston Paris phenomenon, 62
Gauchat, L., 62
Geeraerts, D., 2, 135
Geis, M., 110, 125126
Gerritsen, M., 173
Ghnter, L., 157
Gibson, J.J., 93
Giles, H., 150
Gilliron, J.L., 74, 102
Gilman, A., 111112, 116
Givn, T., 6, 106, 188
glottochronology, 103
Goddard, C., 243, 261
Goddard, Y., 105
Goffman, E., 164165
Golato, A., 249, 253
Goodman, M., 53
Goodwin, M.H., 257
Government & Binding
theory, 206
grammaticalization, 106
Great Vowel Shift, 102
Green, D., 31
Green, Richard E., 82
Greenberg, J.H., 234235
Gregory, M., 226
Grice, H.P., 123, 125
Grices Quantity maxim, 125
Grosjean, F., 19, 21, 23, 3031
Guattari, F., 159160, 167
Gumperz, J., 40, 206, 225, 230
Gumperz, J.J., 78
Gnthner, S., 8
Guy, G.R., 196
H
Haas, M.R., 38
Habermas, J., 88
Haiman, J., 174175
Hall, J., 69
Hall, R.A., Jr., 48, 51, 53
Halliday, M.A.K., 155156,
159162, 164, 221222, 226,
228, 232
Halliday, W.J., 179
Hancock, I., 54

Index

Handler, J.S., 49
Hannah, J., 248
Hansen, L., 40
Harris, A., 182, 208
Harris, A.C., 7
Harris, J., 177
Hartford, B., 152
Haspelmath, M., 6, 18
Haugen, E., 18, 2025, 39, 61,
63, 207
Haugh, M., 250, 261
Hauser, M.D., 8182
Hawaiian Creole, 46
Hawkins, J.A., 236
heavy borrowing, 209
Heidelberger
Forschungsgruppe, 229
Heine, B., 6, 39, 106, 134138,
182, 187, 192, 202, 204
Henrich, J., 86
Herman, E.S., 167
Herring, S.C., 2
Herzog, M., 173
Herzog, M.I., 17
Hewes, G.W., 88
Hickey, R., 3, 5, 1112, 117,
171198
hierarchies of borrowability, 25
Hill, B., 249
Hill, K.C., 54
Himmelmann, N.P., 8
Hinnenkamp, V, 226
Hiri Motu, 47
historical linguistics (HL), 100
historical pragmatics, 110, 241
historical pragmatics, data
problems, 111113
historical pragmatics, origins
of, 111
Hjelmslev, L., 50
HL. see historical linguistics
HL and generative
grammar, 103104
HL in pre-generative
work, 100103
Hofstede, G.H., 261
Holm, J., 46, 48, 5152, 55, 210
Holmes, J., 241, 246, 255, 257
homonymic conflict, 74
Hope, J., 117
Hopper, P., 6
Hopper, P.J., 106107, 126,
135, 188
Horn, L.R., 124125, 127, 129130

House, J., 144, 152, 243

House, R.J., 261
HP. see historical pragmatics
Huang, Y., 240241
Hudson, R., 61
Hudson, R.A., 231
Hundt, M., 192
Hymes, D.H., 241, 246, 255, 257
hyponym, 130
I
Ihalainen, O., 182
Immelmann, K., 88
implicature, metonymy and
merger, 134135
implicature and language
change, 123
implicature and metaphor,
131134
imposition, 20
indigenized varieties, 50
Inhibitory Control Model, 31
insertion, 205
instrumentality in higher
mammals and man, 8891
integration of borrowings,
2324
integrative dialectology, 247
intercultural
communication, 230
intercultural pragmatics, 242
Intercultural Pragmatics, 256
interference, 20, 207
interlanguage, 141
interlanguage pragmatics,
141, 243
Interlanguage pragmatics,
research methods, 152
inter-lingual variation, 249
intertwining, 212
intimate borrowing, 21
intra-lingual variation, 810, 249
Itkonen, E., 108
J
Jaberg, K., 76
Jackendoff, R., 82
Jacobs, A., 110
Jahr, E.H., 68
Jake, J., 207
Jakobson, R., 38, 8384, 102
Janda, R.D., 179
Janney, R., 243
Janney, R.W., 150

Jansen, T., 187

Jargon, 155
jargon, varieties of, 165168
Jautz, S., 257259
Jeffers, R.J., 3
Jespersen, O., 49
Johanson, L., 20
Johnson, E., 69
Johnson, G., 52
Johnson, M., 138
Johnstone, B., 262
Joseph, B., 179, 211
Journal of Historical
Pragmatics, 112
Jucker, A., 187
Jucker, A.H., 110118, 241, 244,
246, 252253, 259
Judd, E., 153
K
Kachru, B., 248
Kachru, B.B., 40
Kdr, D.Z., 117
Kallen, J.L., 255, 258259, 261
Kasper, G., 141153, 249250,
252253
Kaufman, T., 17, 1920, 24,
2627, 39, 55, 108, 195, 208
Keesing, R.M., 52
Keller, R., 108
Kempson, R., 130
Kenstowicz, M., 172
Kerswill, P., 41
Kim, H., 179
King, R., 172
Kinship terminology, 216
Kiparsky, P., 104
Kisseberth, C., 172
Kittay, E., 186
Klimov, G.A.K., 194
Kloss, H., 62
Kluge, F., 166
Koch, P., 111112
Kohnen, Th., 115117
koins, 50
Knig, E., 126127, 133135, 137
Koptjevskaja-Tamm, M., 38
Kopytko, R., 117
Koutsoudas, A., 172
Kouwenberg, S., 5455
Krapp, G.P., 52
Kretzschmar, W.A., 69
Kristiansen, T., 40
Kroch, A., 185

271

272 Index
Kryk-Kastovsky, B., 113
Kurath, H., 52, 66
Kuteva, T., 39, 182, 194
Kyt, M., 112, 113
L
Labov, W., 17, 55, 61, 6364,
6970, 105, 172173, 177178,
191, 231, 247, 251
Lafage, S., 167
Lahiri, A., 182
Lakoff, G., 92, 138
Lalla, B., 49
Langacker, R., 106, 108
Langacker, R.W., 187
language and dialects, 6162
Language as social semiotic, 221
language bioprogram
hypothesis, 53
language change, analogy, 190
language change, comparative
method, 189190
language change, data collection
methods, 192
language change, genre variation
and stylistics, 192
language change,
grammaticalization,
192193
language change, internal
reconstruction, 190
language change, morphological
change, 181182
language change, pathways of
change, 192
language change, phonological
change, 180181
language change, sociolinguistic
investigations, 190191
language change, syntactic
change, 182183
language change, typological
perspective, 193194
language change issues, ebb and
flow, 179180
language change issues, iconicity
and indexicality, 174175
language change issues, internal
and external factors, 173
language change issues,
markedness and
naturalness, 175176
language change issues, mergers
and distinctions, 177178

language change issues, possible

changes, 178
language change issues,
simplicity and
symmetry, 174
language change issues,
telic changes and
epiphenomena, 176177
language change issues,
unidirectionality of
change, 178179
language contact, 203
language mixing in
conversation, 205207
language modes, 31
Languages in contact, 203
language teaching, 151152
language universals and
typology, 234
langue familire, 225
langue populaire, 225
Lanza, E., 41, 205
Lappin, S., 187
Larsen-Freeman, D., 40, 153
Lass, R., 173, 179, 191, 195
Lavandera, B., 251
leaf sponging, 89
Lebsanft, F., 111, 115
lect, 63
Leech, G., 192, 240242
Lefebvre, C., 52, 55
Lehiste, I., 3, 38, 207
Lehmann, Ch., 67
Lehmann, W.P., 194
Leino, J., 7
Lenker, U., 113
Lerud, T., 69
Levinson, S.C., 111, 116, 124126,
129, 139, 237, 241, 250, 258
lexical borrowing, 1819
lexical diffusion, 77
lexical interference, 23
lexicon:grammar-split, 211
lexicostatistics, 103
lexifiers, 47
life-space, 93
Lightfoot, D., 104, 172, 184185
Lillo-Martin, D., 23
Limerick Corpus of Irish
English (LCIE), 253, 259
lingua franca
communication, 242243
linguistic action, production
of, 143144

linguistic atlases, 79
Linguistic Atlas of the Gulf States
(LAGS), 67
linguistic-communicative
registers, 222
linguistic pragmatics, 230231
linguistic reconstruction,
215220
linguistic register forms, 224
linguistic typology, 234237
linguistic variables, 64,
250251
Lipka, L., 135
Listen, P., 117
Li Wei, 40
loan blends, 22
loanshifts, 2223
loanwords, 22
Locher, M., 241, 245
Lombroso, C., 158
Longobardi, G., 182
Ltscher, A., 111, 115
Louis, P., 167
Luckmann, T., 164
Lumsden, J.S., 52
Luraghi, S., 2
M
Macaulay, R., 6170,
256258, 260
Macaulay, R.K.S., 65
Mac Cormac, E.R., 186
Maeschler, Y., 28
Mair, C., 192
Manessy, G., 55
markedness and
naturalness, 175
Markey, T.L., 46
Marples, M., 167
Mrquez Reiter, R., 257
Martinet, A., 64, 68, 102
Martn Rojo, L., 155168
Maschler, Y., 207
Masica, C., 211
Masica, C.P., 38
Mather, J.Y., 66
Matras, Y., 19, 25, 2728, 41,
203212
Matrix Language Frame
model, 206
Mausser, O., 166
Mazzon, G., 118
Mboudjeke, J.-G., 29
McColl Millar, R., 171

Index

McDavid, R.I., 61
McDavid, V., 52, 61, 69
McDavid Jr., R., 52
McIntosh, A., 221, 228
McLaughlin, M.L., 245
McMahon, A., 171
McWhorter, J., 210
McWhorter, J.H., 50
Meechan, M., 21, 23
Meeuwis, M., 12, 28, 3643,
108, 206
Meillet, A., 6, 106
Meisel, J.M., 205, 212
Melanesian pidgins, 52
Melchers, G., 249
Merle, P., 167
mesolects, 55
Mesthrie, R., 196
metalingual function, 83
metatypy, 208
Metcalf, G.J., 111, 117
metonymy, 135
Michael, S., 156
Miller, C., 18, 24, 27,
2930, 206
Miller, G.D., 24, 27
Milroy, J., 20, 171, 177, 191,
247, 249
Milroy, L., 64, 177, 191, 247
Mixed Languages, 211
mixed lects, 63
Modern New Synthesis, 83
Moravcsik, E., 208
Moravcsik, E.A., 18, 25
Mous M., 27, 41, 211212
Moyer, M., 28
Mufwene, S., 13, 4655
Mufwene, S.S., 46
Mhlhusler, P., 4748, 5455
Muhr, R., 254, 256, 258
Muntzel, M.C., 42
Murphy, B., 258259
Mutz, K., 8
Muysken, P., 1819, 2125, 27,
30, 5455, 205206
Myers-Scotton, C., 205206, 212
N
Nabrings, K., 228, 231
Nnny, M., 174
nasalization, 190
native speaker non-native
speaker
communication, 242

Natural Semantic
Metalanguage, 243
Nelde, P.H., 204
Neodarwinism, 83
Neo-Gricean theories of
communication., 123
Nevala, M., 112
Nevalainen, T., 2, 112, 118
Newmeyer, F.J., 183
Niceforo, A., 157159
Nichols, J., 103, 194
niveau courant, 224
niveau familier, 224
niveaux de langue, 224225
Noble, W., 88, 91
nonce borrowings, 23
Norde, M., 179
O
OBarr, W., 167
Objartel, G., 167
Obler, L.K., 204
Oesterreicher, W., 111112
OKeeffe, A., 256, 259
Oldendorp, C.G.A., 48
Oleksy, W., 242
Olshtain, E., 142143
Onodera, N.O., 114
Orton, H., 65, 179
Osgood, C., 204
Osthoff, H., 215
stman, J-O., 1112, 3643, 81
P
Pagliuca, W., 106, 131, 193
Pakkala-Weckstrm, M., 116
Palmireno, L., 160
Paolillo, J., 38
Paradis, M., 204
Parry, J.H., 166
participant-related
code-switching, 206
Patrick, P., 55
Patridge, E., 166
pattern knowledge, 231
Paul, H., 172
Peirce, C.S., 81, 94
Person, R.R., Jr., 114
Pfaff, C., 19
phatic function, 83
Phillips, B., 186
Pichler, H., 256258, 260
pidgin, 46
Pidginization Index, 53

Piirainen-Marsh, A., 150

Pinker, S., 82
Pino, C., 39
Pintzuk, S., 182, 185
Piveteau, J., 90
Placencia, M.E., 254, 257258, 260
Plag, I., 49
Plevoets, K., 248, 256, 260
Plmacher, Martina, 84
pluricentric languages, 248
poetic function, 84
politeness, 116118
Poplack, S., 1819, 21, 2324,
2930, 41, 49, 54, 206207
Porter, R., 166, 167
Posner, R., 50, 53
post-colonial pragmatics, 243
post-creole continuum, 49
post-Gricean theories of
communication., 123
Poussa, P., 38
pragmalinguistics, 242
pragmatic analysis, levels
of, 244246
pragmatic borrowing, 19, 29
pragmatic change, 187188
pragmatic comprehension,
assessment of
politeness, 142
pragmatic comprehension,
nonliteral utterances,
141142
pragmatic comprehension,
sociopragmatic
assessment, 142143
pragmatic development,
cross-sectional
studies, 144145
pragmatic development,
longitudinal studies, 145
pragmatic development,
theoretical accounts
of, 145146
Pragmatic functions and
hierarchy of signs, 84
pragmatic norms, 150151
pragmatics, 81
pragmatics, place in, 241244
pragmatics, scope of, 240241
pragmatic strengthening, 6
pragmatic transfer,
definition, 146
pragmatic transfer, negative
transfer, 147148

273

274 Index
pragmatic transfer, positive
transfer, 146147
pragmatic transfer,
transferability, 148149
pragmatic universals, 77
pragmatic variable, 250252
pragmatic variation, 249250
Pragmatic variation: The
interplay of micro-social and
macro-social factors, 255
Pragmatism, 81
Preston, D.R., 68
Primaz, L., 167
Prince, E., 39
Prince, E.F., 19, 2829
Principle of
Informativeness, 125
Principle of Relevance, 129
Proto-Indo-European, 218
proto-languages, 218
Proto-Romance, 219
prototypical implicature, 123
Proto-X, 216
psycholinguistic approaches to
borrowing, 3032
Pujolar, J., 167
Q
Q-based narrowing, 130
Q-inference, 125
Q Principle, 125
Quantitative approaches to
borrowing, 2930
Queen, R.M., 19
Quiatt, D., 88
R
Ramat, P., 38
Raukko, J., 38, 43
Raumolin-Brunberg, H., 118
R-based broadening, 131
Redeker, G., 187
Reetz-Kurashige, A., 40
referential function, 83
regional variation, 247
register, history of term,
221224
register, systematization of
term, 224228
Rehbein, J., 230231
Reid, T.B.W., 221
Relevance Theory, 129
relexification hypothesis, 52
representation, 83

Reynolds, P.C., 90
Reynolds, V., 88
Rickford, J.R., 49, 55
R-inference, 125
Ringbom, H., 4041
Robert LArgenton, F., 166
Roberts, I., 182, 184185
Roberts, S.J., 47, 54
Roche, J., 223, 230
Roeding, J.H., 166
Roitblat, H., 142
Roma, E., 38
Romaine, S., 1819, 55,
106, 198
roofless dialects, 62
Rosenbach, A., 179
Ross, M., 188, 208, 212
Rost, M., 229
R Principle, 125
Ruhlen, M., 105
S
Sacks, H., 246, 249
Sainan, L., 157
Sakel, J., 2728
Salillas, R., 157158
Salmons, J., 19, 28, 207, 210
Sanders, C., 224, 231
Sandler, W., 23
Sandy, H., 40
Sankoff, D., 18, 24, 2930,
52, 206
Sankoff, G., 52, 54
Sanmartn, J., 156, 163
Sastre, A., 161
scalar implicature, 124
Schegloff, E.A., 249
Schiffrin, D., 246
Schilling-Estes, N., 171, 245247,
255, 258, 262
Schleicher, A., 51, 101
Schlieben-Lange, B., 111,
239, 255
Schmid, H.-J., 187
Schmidt, J., 101, 144146
Schmidt, R., 101, 144146
Schneider, E.W., 49, 53, 65
Schneider, K.P., 239262
Schlmberger, U., 257259
Schrott, A., 111, 115
Schuchardt, H., 49
Schwenter, S.A., 193
Sebba, M., 41
Seidlhofer, B., 40

selective value of
communication and
symbolic behavior, 8687
Selinker, L., 141, 243
Selting, 226, 231
semantic change, 186187
semantic loan, 23
semi-creoles, 50
semiotic signs, 83
Shaw, P., 249
Siegel, J., 54
Silva-Corvaln, C., 207
Simon, H.J., 117
Singh, I., 171
Singh, R., 206
Singler, J.V., 5354
Singy, P., 166
Skewis, M., 117
Smith, N., 54, 192
Soares, C., 31
social communication, 85
social dialects, 64
Social Matrix (1991), 69
social structure and language
structure, 227
social variation, 247
societal multilingualism,
203204
sociopragmatics, 242
socio-situational variations, 224
Sll, L., 111
Somolinos, A.R., 114
Spanish-English
bilingualism, 206
speech act modification, 245
speech act realization, 245
speech acts, 114116
Speitel, H.-H., 66
Sperber, D., 129
Sprachatlas des deutschen
Reichs, 79
Sprachbund/Sprachbnde, 38,
197198, 210
Sridhar, K.K., 243
Stein, D., 173
Stern, G., 135
Stewart, W.A., 62
Stolz, C., 28, 209, 212
Stolz, T., 28, 209, 212
Strevens, P., 221, 228
structural borrowing, 19, 26
substrate hypothesis, 5152
superstrate hypothesis, 5253
superstratum language, 41

Index

Sweetser, E., 107
Sweetser, E.E., 138
Sylvain, S., 5152
symmetrical
communication, 249
synergy, 42
syntactic borrowing, 26
T
Taavitsainen, I., 111, 115117, 241,
244, 252
Tadmor, U., 18
Tagliamonte, S., 49, 54
Takahashi, S., 142, 148
Talmy, L., 138
Taylor, J., 136
Tembrok, G., 88
tenor (discourse style), 222
Terkourafi, M., 251252
The General Index (1988), 68
The Handbook (1986), 68
The Linguistic Atlas of the Middle
and South Atlantic States
(LAMSAS), 6869
The Phonological Atlas of North
America (TELSUR), 6870
The Pragmatics of Irish
English, 255
The Regional Matrix (1990), 69
The Regional Pattern (1991), 69
The Social Pattern (1992), 69
The Technical Index (1989), 69
Thom, R., 84
Thomas, J., 149, 242
Thomason, S.G., 17, 1920,
24, 2627, 39, 42, 55, 108,
195196, 208209, 211212
Thomsen, N., 4
Tieken-Boon van Ostade, I., 112
Tomasello, M., 87
tool manufacturing, 9192
Torrione, M., 160
Tottie, G., 262
transference, 20
Trask, R.L., 171, 182, 190
Traugott, E., 126, 135, 171, 192
Treffers-Daller, J., 12, 1732
triad of sign functions, 8384
Trier, J., 102
Trosborg, A., 153, 243
Trudgill, P., 6163, 65, 67, 171,
179, 191, 248
Tsoulas, G., 182
Turner, L.D., 51

U
Ullmann, S., 107
Ungerer, F., 187
Unitary Language System
Hypothesis, 204
universal grammar, principles
and parameters
model, 184186
universalist hypotheses, 53
Ureland, S., 38
V
Valds, G., 39
Valdman, A., 54
Valkhoff, M.F., 48
Valkonen, P., 116
Van Coetsem, F., 20, 24, 39
van der Auwera, J., 179
Van linden, A., 6
variation, 1
variational pragmatics, 239240
Variational pragmatics:
A focus on region, age and
gender, 255
Variational Pragmatics: A focus
on regional varieties in
pluricentric languages, 255
Variational pragmatics:
Cross-cultural
approaches, 255
variational pragmatics,
data, 253254
variational pragmatics,
history, 254256
variational pragmatics,
methodological
principles, 252253
variational pragmatics, research
trends, 256260
variational pragmatics vs.
historical pragmatics,
243244
variation and identity, 246249
variation linguistics, 231232
variety, 6263
Vennemann, T., 106, 236
vernacular, 6364
Verschueren, J., 81, 237, 240
Verspoor, M., 187
Vidocq, E., 157
Vincent, N., 182
Vinson, J., 51
Vlker, H., 111
Volterra, V., 204

W
Wade-Lewis, M., 52
Wagner, H., 198
Wagner, M.L., 157
Wang, S., 186
Wang, W., 186
Warga, M., 258259
Warner, A., 182
Warren, B., 132, 138
Watts, R., 241
Watts, R.J., 116
Wawrzyniak, U., 24
Weinreich, U., 1718, 2021, 25,
27, 39, 173, 203204, 210
Wellington Corpus of
SpokenNew Zealand
English, 253
Weltens, B., 67
Weydt, H., 239, 255
Whitney, W.D., 25
Widdowson, J., 65
Wiemann, J.M., 150
Wiemer B., 7
Wierzbicka, A., 187, 237, 243,
249250, 261
Wildgen, W., 8197
Wildner-Bassett, M., 151
Williams, S.W., 52
Wilson, D., 129
Winford, D., 27, 39
Wischer, I., 179
Wolfram, W., 65, 245247, 255,
258, 262
Wolfson, N., 153, 250
Woolard, K.A., 42
Wunderli, P., 232
Y
Yates, F.A., 93
Yoffe, M., 167
Yule, G., 246
Z
Zwicky, A., 110, 125126
Zwicky, A.M., 110, 125126

275

In the series Handbook of Pragmatics Highlights the following titles have been published thus far or
are scheduled for publication:
10 SBIS, Marina, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Philosophical Perspectives. ca.250pp.
Expected Forthcoming
9 STMAN, Jan-Ola and Jef VERSCHUEREN (eds.): Pragmatics in Practice. ca.250pp. Expected Forthcoming
8 LEDIN, Per, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Discursive Pragmatics. ca.250pp. Expected
Forthcoming
7 JASPERS, Jrgen, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Society and Language Use. 2010. xiii,324pp.
6 FRIED, Mirjam, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Variation and Change. Pragmatic perspectives.
2010. x,275pp.
5 BRISARD, Frank, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Grammar, Meaning and Pragmatics. 2009.
xiii,308pp.
4 DHONDT, Sigurd, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): The Pragmatics of Interaction. 2009.
xiii,262pp.
3 SANDRA, Dominiek, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Cognition and Pragmatics. 2009.
xvii,399pp.
2 SENFT, Gunter, Jan-Ola STMAN and Jef VERSCHUEREN (eds.): Culture and Language Use. 2009. xiii,280pp.
1 VERSCHUEREN, Jef and Jan-Ola STMAN (eds.): Key Notions for Pragmatics. 2009. xiii,253pp.