PDF

CONTEMPORARY APPLIED LINGUISTICS
VOLUME 2 LINGUISTICS FOR THE REAL WORLD

Continuum Contemporary Studies in Linguistics
Continuum Contemporary Studies in Linguistics series presents a snapshot of the
current research being undertaken in the core areas of linguistics. Written by
internationally renowned linguists, the volumes provide a selection of the best
scholarship in each area. Each of the chapters appears on the basis of its impor-
tance to the field, but also with regards to its wider significance either in terms of
methodology, practical application or conclusions. The result is a stimulating con-
temporary snapshot of the field and a vibrant reader for each of the areas covered
in the series.
Contemporary Stylistics
Edited by Marina Lambrou and Peter Stockwell
Contemporary Corpus Linguistics

Edited by Paul Baker
Contemporary Applied Linguistics Vols 1 and 2

Edited by Li Wei and Vivian Cook
Contemporary Applied
Linguistics
Edited by
Li Wei and Vivian Cook
Volume 2 Linguistics for the Real World
Edited by
Li Wei
Continuum International Publishing Group
The Tower Building 80 Maiden Lane, Suite 704
11 York Road New York
London SE1 7NX NY 10038
© Li Wei, Vivian Cook and Contributors 2009
All rights reserved. No part of this publication may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including photocopying, recording, or
any information storage or retrieval system, without prior permission in writing from the
publishers.
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.
ISBN: 978-0-8264-9681-2 (Hardback)
Library of Congress Cataloging-in-Publication Data

The Publisher has applied for CIP data.
Typeset by Newgen Imaging Systems Pvt Ltd, Chennai, India

Printed and bound in Great Britain by the MPG Books Group
Contents
Notes on Contributors vii

Acknowledgements xi
Introduction: Linguistics for the Real World 1

Chapter 1 Globalization, Multilingualism, and Gender: 10
Looking into the Future
Ingrid Piller and Aneta Pavlenko
Chapter 2 Language and Economy 28
Florian Coulmas
Chapter 3 Linguistic Diversity and Poverty: Many Languages and 46
Poor People in a Globalizing World
Suzanne Romaine
Chapter 4 Religious Language Management 65
Bernard Spolsky
Chapter 5 Language and Culture 83
Nick Enfield
Chapter 6 Four-Borne Discourses: Towards Language as 98
a Multi-Dimensional City
Gu Yueguo
Chapter 7 Discourse in Organizations and Workplaces 122
Britt-Louise Gunnarsson
Chapter 8 Political Discourse and Translation 142
Christina Schäffner
Chapter 9 Language and the Law 164
John Gibbons
Chapter 10 Neurolinguistics and the Non-monolingual Brain 184
Marjorie Lorch
vi Contents
Chapter 11 Clinical Phonology 202

Martin J. Ball and Nicole Müller
Chapter 12 Sign Linguistics, Sign Language Learning and 223
Sign Bilingualism
Gary Morgan and Bencie Woll
Personal Name Index 243

Subject Index 244
Notes on Contributors
Martin Ball, Ph.D., mjb0372@louisiana.edu, is Hawthorne-BoRSF Endowed Pro-

fessor at the University of Louisiana at Lafayette. He is co-editor of the journal
Clinical Linguistics and Phonetics. His research interests are sociolinguistics, clinical
phonetics and phonology, and the linguistics of Welsh. His most recent book is
Handbook of Clinical Linguistics (co-edited with Mick Perkins, Nicole Müller and
Sara Howard. Blackwell, 2008).
Vivian Cook, Ph.D., vivian.cook@ncl.ac.uk, is Professor of Applied Linguistics at

Newcastle University, UK. He is interested in the multi-competence approach to
SLA research and writing systems. He is joint editor of the journal Writing Systems
Research (OUP).
Florian Coulmas, Ph.D., coulmas@dijtokyo.org, is Director of the German Institute

for Japanese Studies, Tokyo and Professor of Japanese Studies, University Duisburg-
Essen, on leave of absence. He has taught and done research in various environ-
ments, including The National Language Research Institute, Tokyo, Georgetown
University, Washington, DC, USA, and Chuo University, Tokyo. His most recent
books in linguistics are Language Regimes in Transformation (ed.), Mouton de Gruyter
2007, and Sociolinguistics: The Study of Speakers’ Choices, Cambridge University Press
2005.
Nick Enfield, Ph.D., nick.enfield@mpi.nl, is a senior staff scientist in the Language

and Cognition Group, Max Planck Institute for Psycholinguistics, Nijmegen. His
research interests include language, culture, cognition and social interaction. His
books (edited and authored) include Ethnosyntax (2002), Linguistic Epidemiology
(2003), Roots of Human Sociality (with SC Levinson, 2006), Person Reference in Interac-
tion (with T Stivers, 2007), A Grammar of Lao (2007) and The Anatomy of Meaning
(2009).
John Gibbons, Ph.D., johgibbo@hotmail.com, was Professor of Linguistics at

Hong Kong Baptist University until retirement. He had previously worked in the
University of Sydney, Australia and the University of Hong Kong. His main research
interests are language and the law, and bilingualism. His publications include
Language and the Law (Longman, 1994) and Forensic Linguistics (Blackwell, 2003).
viii Notes on Contributors
Gu Yueguo, Ph.D., yueguo.gu@gmail.com, is Research Professor of Linguistics at

the Chinese Academy of Social Sciences. His research interests include: Corpus
linguistics, pragmatics, discourse analysis and online education. His recent publi-
cations included ‘Multimodal Text Analysis: A Corpus Linguistics Approach’(Text
and Talk, 2006), Using the Computer in ELT: Technology, Practice, and Theory (2006,
Beijing: FLTRP) and Exploring Online Education (2007, Beijing: FLTRP).
Britt-Louise Gunnarsson, Ph.D., britt-louise.gunnarsson@nordiska.uu.se, is Profes-

sor of Modern Swedish and Sociolinguistics at Uppsala University. Her research
interests are sociolinguistics, text linguistics, discourse analysis and professional
communication. Her recent publications include the single-authored monograph
Professional Discourse (Continuum, 2009) and chapters published in Encyclopedia of
Language and Education (2008), Sage Benchmarks in Discourse Studies (2007), Hand-
book of Pragmatics (2006) and Encyclopedia of Language and Linguistics (2005).
Li Wei, Ph.D., li.wei@bbk.ac.uk, is Professor of Applied Linguistics at Birkbeck

College, University of London. His research interests include bilingual and inter-
cultural pragmatics. He is Principal Editor of the International Journal of Bilingual-
ism (Sage). His recent publications include the Blackwell Guide to Research Methods
in Bilingualism and Multilingualism (with Melissa Moyer, 2007) and the Handbook of
Multilingual and Multilingual Communication (with Peter Auer, 2008, Mouton de
Gruyter).
Marjorie Lorch, Ph.D., m.lorch@bbk.ac.uk, is Professor of Neurolinguistics at

Birkbeck College, University of London. Her research investigates the function of
language as a human biological endowment through a range of issues including
neurological disorder, development and aging, bilingual and cross-linguistic issues,
communication systems and literacy. Her publications engage a large international
audience bridging applied linguistics to medical practitioners, psychologists, edu-
cators and historians.
Gary Morgan, Ph.D., g.morgan@city.ac.uk, is Professor of Psychology at City Uni-

versity London and Deputy Director of DCAL at UCL. His research interests
include: sign languages, bilingualism and deaf children. His recent publications
include papers in Cognitive Development, First Language, Applied Psycholinguistics and
Journal of Child Language. With colleagues he is just finishing a book on sign lan-
guage learning in the polyglot Savant Christopher – Smith, N., Tsimpli, I. M., Morgan,
G. and Woll, B. Signs of the Savant. Cambridge University Press.
Nicole Müller, Ph.D., nmueller@louisiana.edu, is Professor of Communicative

Disorders at the University of Louisiana at Lafayette. She has recently co-edited the
Handbook of Clinical Linguistics, and co-authored Phonetics for Communication Disorders,
Phonology for Communication Disorders, and Approaches to Discourse in Dementia. She is
co-editor of Clinical Linguistics and Phonetics.
Notes on Contributors ix
Aneta Pavlenko, Ph.D., apavlenk@temple.edu, is Professor of TESOL at Temple

University. Her research involves psycholinguistics and sociolinguistics of bilin-
gualism. Her recent books include Emotions and Multilingualism (Cambridge Uni-
versity Press, 2005), Bilingual Minds (Multilingual Matters, 2006), Crosslinguistic
Influence in Language and Cognition (with Scott Jarvis, Routledge, 2008), and Multi-
lingualism in Post-Soviet Countries (Multilingual Matters, 2008).
Ingrid Piller, Ph.D., ingrid.piller@gmail.com, is the founding director of the

Center for Bilingualism and Bilingual Education in the Arab World at Zayed Uni-
versity in the United Arab Emirates. Her research is in the sociolinguistics of lan-
guage learning and intercultural communication. Her books include Bilingual
Couples Talk (Benjamins, 2002) and she is currently working on a textbook on Inter-
cultural Communication for Edinburgh University Press.
Suzanne Romaine, D.Phil., suzanne.romaine@ling-phil.ox.ac.uk has been Merton

Professor of English Language at the University of Oxford since 1984. She has
received honorary doctorates from the University of Uppsala and the University of
Tromsø. She is co-author of Vanishing Voices. The Extinction of the World’s Languages,
which won the British Association of Applied Linguistics book of the year prize in
2001.
Christina Schäffner, Ph.D., c.schaeffner@aston.ac.uk, is Professor of Translation

Studies at Aston University, Birmingham. Her research interests are Translation
and Politics, Political Discourse Analysis; Metaphor and Translation, and Transla-
tion Didactics. Her recent publications include Translation Research and Interpreting
Research: Traditions, Gaps and Synergies (ed., 2004), ‘Political Discourse Analysis
from the Point of View of Translation Studies’, Journal of Language and Politics,
3/2004, ‘Metaphor and Translation: Some Implications of a Cognitive Approach,
Journal of Pragmatics, 2004.
Bernard Spolsky, Ph.D., bspolsky@gmail.com, is professor emeritus in the Depart-

ment of English at Bar-Ilan University Israel. His most recent research interest is
language policy and language management. His books include Language Policy
(Cambridge University Press in 2004), The Handbook of Educational Linguistics
(edited with Francis Hult and published by Blackwell, 2008) and Language Manage-
ment (Cambridge, 2009).
Bencie Woll, Ph.D., b.woll@ucl.ac.uk, is Professor of Sign Language and Deaf

Studies, and Director of the Deafness Cognition and Language Research Centre
at UCL. Her research interests embrace a wide range of topics related to sign
language. These include the linguistics of British Sign Language (BSL) and
other sign languages, the history and sociolinguistics of BSL and the Deaf commu-
nity, the development of BSL in young children, and sign language and the
brain. Her publications include The Linguistics of BSL: An Introduction (with Rachel
x Notes on Contributors
Sutton-Spence), which was the winner of the 1999 Deaf Nation Award and 2000
British Association of Applied Linguistics Book Prize, and she has co-edited with
Roland Pfau and Markus Steinbach Sign Language: An International Handbook (in
preparation) and Sign Language Acquisition (with Anne Baker) (in press).
Acknowledgements
We are indebted to the contributors for their work in bringing these volumes into
being; we as editors have learnt so much both from their previous work and from
their contributions here; we hope the process was as rewarding for them as it was
for us. Vivian Cook would also like to thank his TESOL colleagues at Newcastle
University and Jean-Marc Dewaele for help and advice on particular aspects of
Volume 1. His share would never have been completed without the musical sup-
port of Eric Dolphy, Humphrey Lyttelton and Mike Osborne, whose memories live
on in their recordings. Li Wei would like to thank Joshua Borden and Zhu Hua for
their help with editing some of the chapters. Finally, we are grateful to Continuum
for the opportunity of working together on this project.
This page intentionally left blank
Introduction: Linguistics
for the Real World
1 General Background
Since the days of Pit Corder, the founding father of British applied linguistics in
the 1950s, the discipline of applied linguistics has been usually described as ‘The
theoretical and empirical investigation of real-world problems in which language
is a central issue’ (Brumfit, 1995: 27). Similarly, the members of the American
Association of Applied Linguistics (AAL) ‘promote principled approaches to lan-
guage-related concerns’. The International Association of Applied Linguistics
(AILA) proclaims:
applied linguistics is an interdisciplinary field of research and practice dealing

with practical problems of language and communication that can be identified,
analysed or solved by applying available theories, methods or results of Linguis-
tics or by developing new theoretical and methodological frameworks in
Linguistics to work on these problems.
The AILA definition is both broader in including more areas and narrower in
relating applied linguistics to linguistics proper. If you have a problem with lan-
guage, send for an applied linguist.
The broad definition of applied linguistics as problem-solving was certainly true
in its early days. Definitions of applied linguistics now are more like lists of the
areas that make it up. The Cambridge AILA 1969 Congress encompassed first lan-
guage acquisition, computational linguistics, forensic linguistics, speech therapy,
neurolinguistics, second language acquisition research, and a host more. Gradu-
ally many areas have declared unilateral independence from applied linguistics;
first language acquisition research soon disappeared from the fold to found its
own organizations, conferences and journals, as did much second language acqui-
sition research slightly later. Applied linguistics gatherings these days are far less
inclusive, though there is a growth in the Research Networks such as Multilingual-
ism: Acquisition and Use. The AILA Congress in 2008 had 9 papers on first lan-
guage acquisition compared with 161 on second language acquisition and 138 on
foreign language teaching; computational linguistics and forensic linguistics were
no longer on the programme, though new areas like multilingualism have been
introduced. Professional organizations for applied linguistics are now more like
umbrella organizations, on the lines of the British Association in science, that meet
2 Li Wei and Vivian Cook
occasionally to bring together people whose main academic life takes place within
more specialist organizations; most second language acquisition researchers
for instance tend to go to conferences of the European Second Language Associa-
tion (EUROSLA), the International Symposia on Bilingualism (ISB), Generative
Approaches to Second Language Acquisition (GASLA), or the International Asso-
ciation for Multilingualism, not to conferences named applied linguistics. Profes-
sional applied linguistics is now a fairly restricted area. Most practitioners probably
style themselves primarily as SLA researchers, discourse analysts and the like,
rather than seeing applied linguistics as their major avocation. Journals too reflect
this tendency with say the International Journal of Applied Linguistics showing the
same kind of agenda as the AILA congress, while Language Learning, originally
an applied linguistics journal, is now primarily concerned with psychological
approaches to second language acquisition, having dropped applied linguistics
from its subtitle.
The term ‘problem’ does, however, raise issues of its own. In one sense it means
a research question posed in a particular discipline; in another sense it is some-
thing that has gone wrong which can be solved. Talking about the problem of
multilingualism, say, is ambiguous between defining it as a research area and claim-
ing that it is in some way defective. Calling areas problems fosters the attitude that
there is something wrong with them. Bilingualism is no more intrinsically a prob-
lem to be solved than is monolingualism. Applied linguists have to be clear that
they are solving problems within an area of language acquisition or use, not regard-
ing the area itself as a problem except in the research question sense. Language
teaching is not itself a problem to be solved; it may nevertheless raise problems
that applied linguists can resolve.
A perpetual controversy has surrounded the relationship of linguistics to applied
linguistics. Despite AILA’s fond belief that linguistics is the core, many feel linguis-
tics is only one of the contributing disciplines. Applied linguists have explored psy-
chological models such as declarative/procedural memory and emergentism,
mathematical models such as dynamic systems theory or chaos theory, early Soviet
theories of child development such as Vygotsky, French thinkers such as Foucault
and Bourdieu – nothing seems excluded. Contemporary applied linguists feel free
to draw on almost any field of human knowledge; the authors in this book for
instance use ideas from philosophy, education, sociology, feminism, Marxism,
Conversation Analysis and media studies, to take a small sample. Ben Rampton
(1997: 14) described applied linguistics as ‘an open field of interest in language’,
while David Block (2009) calls it ‘an amalgam of research interests’ (p. 2, n.1, Vol.1).
The question is whether applied linguists have the polymathic ability to carry out
such an amalgamation of diverse disciplines, or indeed diverse approaches within
these disciplines, when the disciplines themselves are incapable of making this
synthesis. It seems inherently unsafe or indeed arrogant when the applied linguist
redefines the human mind, human language or language learning to suit the
needs of an applied linguistic problem.
Linguistics nowadays plays a minimal role in applied linguistics whether in terms
of current linguistic theories or descriptive tools. Linguistic theories of the past
20 years are barely mentioned by applied linguists. With the exception of Chomsky
Introduction 3
and to some extent Jackendoff, the theories come from post-modernism, psycho-
logy or sociology rather than linguistics. Indeed some practitioners radiate hostility
towards linguistics, preferring to draw on almost any other area. One cause may be
that the enthusiastic selling of the 1980s generative model by its supporters led to
the view that linguistics has nothing practical to contribute and to a lack of interest
in the many other approaches to linguistics practised today, say the recent develop-
ments in phonetics and phonology.
In a well-regarded book representative of the field called Research Methods in
Applied Linguistics, the author announces ‘The book . . . will not be concerned
with . . . language data, unless it is submitted to non-linguistic analysis’ (Dornyei,
2007: 19). In the west London suburb of Ealing there was a highly successful shop
in the 1960s called the Confiserie Française (French cake shop), which in fact sold
toys. The reason was a clause in its lease that prevented the new owners from
changing the name. If language disappears from applied linguistic research, the
applied linguistics shop is selling toys. It should relabel itself as teaching methodol-
ogy or applied sociology or whichever discipline it uses as its source.
So what problems does applied linguistics solve? If you are worried about your
child’s speech, you are more likely to go to a speech therapist than to an applied
linguist. If your country is torn by civil war between people who use two scripts, you
ask for a United Nations Peacekeeping Force. If you are drafting a new law, you go
to a constitutional lawyer or a civil servant. The problem-solving successes of
applied linguistics have included devising orthographies for languages that have
no written form and inventing simplified languages for mariners; applied linguists
have played a part in EU projects on translation and on linguistic diversity. Most
successes have, however, had to do with language teaching, such as the syllabuses
and methods that swept the world from the 1970s onwards, particularly associated
with the Council of Europe.
At a general level we can draw three implications from this. Needless to say,
these personal interpretations are not necessarily shared by all the contributors.
(1) The applied linguist is a Jack of all trades; real-world language problems can sel-
dom be resolved by looking at a single aspect of language. Since applied lin-
guistics is interdisciplinary, the applied linguist is expected to know a little
about many areas, not only of language, but also of philosophy, sociology,
computer programming, experimental design and many more. In a sense
applied linguists are not only Jack of all trades but also master of none as they
do not require the in-depth knowledge of the specialist so much as the ability
to filter out ideas relevant to their concerns. An applied linguist who only does
syntax or discourse analysis is an applied syntactician or an applied discourse
analyst, not a member of the multidisciplinary applied linguistics profession.
In other words multidisciplinarity applies not just to the discipline as a whole
but also to the individual practitioner.
(2) The applied linguist is a go-between, not an enforcer, a servant, not a master. The
problems that applied linguistics can deal with are complex and multi-facetted.
As consultants to other people, applied linguists can contribute their own
interpretation and advice. But that is all. The client has to weigh in the balance
all the other factors and decide on the solution. Rather than saying ‘You
should follow this way of language teaching,’ the applied linguist’s advice is
‘You could try this way of language teaching and see whether it works for you.’
Alternatively the applied linguist should be responding to problems put
forward by language teachers, not predetermining what the problems are; the
applied linguist is there to serve teachers’ needs, a garage mechanic interpret-
ing the customer’s vague idea of what is wrong with their car and putting it
right, rather than a car designer.
(3) Sheer description of any area of language is not applied linguistics as such but descriptive
linguistics. Some areas concerned with the description of language are regarded
as applied linguistics, others are not. Make a corpus analysis of an area or carry
out a Conversation Analysis and you’re doing applied linguistics; describe chil-
dren’s language or vocabulary and it’s first language acquisition; make a
description of grammar and you’re doing syntax. Overall making a description
is not in itself solving a problem, even if it may contribute to the solution.
Outside language teaching, applied linguists have taken important roles behind
the scenes as advisors to diverse governmental and EU bodies, e.g., Hugo Baetens
Beardsmore’s work with bilingualism. But they have had little impact on public
debate or decision-making for most language problems, the honourable excep-
tions being the work of David Crystal and Debbie Cameron, whom many might not
consider primarily as applied linguists. Problems are not solved by talking about
them at applied linguistics conferences; the solutions have to be taken out into the
world to the language users. Take the political correctness issue of avoiding certain
terms for reasons of sexism, racism and so on. This is based on one interpretation
of the relationship between language and thinking: not having a word means you
can’t have the concept, as George Orwell suggested with Newspeak. Yet applied
linguists have been reluctant to contribute their expertise to this debate, despite
the extensive research into linguistic relativity of the past decade. Public discussion
of language issues is as ill-informed about language as it was 50 years ago at the
dawn of applied linguistics.
A recent theatre piece by the Canadian director Robert Le Page, called Lipsynch,
was crucially concerned with language. The dialogue took place in three languages
with the aid of subtitling running along the front of the stage; it took for granted
the multilingualism of the modern world. The heroine was attempting to recover
the voice of her father who had died when she was young. All she had was a silent
home movie. So she engaged a lip-reader to find out the words, then a lipsynch
actor to read them in alternative voices till she recognized her father’s. This didn’t
work until she herself uttered her father’s words. In another scene an elderly apha-
sic patient delivers a monologue, judging by audience reaction the first time that
most of them had encountered this kind of discourse. At a dinner party, film actors
and agents attempted to converse simultaneously in three languages, to comic effect.
Lipsynch movingly showed the importance of language to people’s lives and the
language problems they encountered.
Introduction 5
As this reminds us, language is at the core of human activity. Applied linguistics
needs to take itself seriously as a central discipline in the language sciences dealing
with real problems. Applied linguistics has the potential to make a difference.
2 Linguistics for the Real World
This volume attempts to reassert the importance of applied linguistics in social life.
It assumes that the unique selling point of applied linguistics that distinguishes it
from the many domains and sub-domains of sociology, economics, politics, law,
management and neuroscience is language. At its core it needs a coherent theory
of language, whether this comes from linguistics or from some other discipline,
a set of rigorous descriptive tools to handle language, and a body of research rele-
vant to language practice.
This is not to say that the language element has to dominate or that a particular
linguistics theory or model has to feature, but it does not count as applied
linguistics
1. if there is no language element. Many of the concerns applied linguists have are also
concerns of sociologists, neuroscientists and other professional researchers.
Crucially, however, applied linguists focus on the role of language in the broad
issues of sociological or neurological concern. Why call it applied linguistics if
it has no language connection?
2. if the language elements are handled without any theory of language. The theory of
language does not need to come from linguistics but might be philosophy,
history, social theory or literary theory. Yet applied linguistics cannot treat lan-
guage as if there were no traditions of language study whatsoever. Nor can the
language elements be based solely on folk ideas from the school tradition of
grammar which would be rather like basing physics on folk beliefs or alchemy.
Indeed, one of the responsibilities of the applied linguist should be to chal-
lenge both the folk notions of language and grammar and the theoretical lin-
guists’ models of how language works.
3. if the research base is neither directly concerned with language issues nor related to them
in a demonstrable way. That is to say, a theory from other disciplines cannot be
applied without a clear chain showing how and why it is relevant. An idea from
mathematical theory, computer simulation or neural network needs to show
its credentials by proving practical solution of real-life language problems (e.g.
how bilingual speakers with aphasia process sentences), not imposing itself by
fiat, by analogy or by sheer computer modelling. This is an area where there is
huge potential for further development, as more and more applied linguists
become attracted by theories and ideas from other disciplines.
Over the last two decades, there have been a number of attempts to reconceptu-
alize applied linguistics as part of social science (e.g. Brumfit, 1997; Rampton,
2000; and most explicitly, Sealey and Carter, 2004). What this means essentially has
been to view language use (including language teaching and learning, language
planning, language testing, etc.) as social practice and empirical analysis of
social practice as social science. Along with some of their sociolinguist colleagues,
applied linguists have borrowed theories and approaches from sociologists such
as Durkheim, Parsons and Levi-Strauss, as well as Bourdieu and Giddens, and
highlighted the differential distributions of linguistic resources in society. One
outcome of such work is an increased awareness among applied linguists that
failure to achieve a certain standard in second language learning, for example, is
not simply a linguistic or psychological matter but a much broader socio-political
issue. Different (varieties of) languages carry different historical and cultural
meanings to different groups of people, resulting in different attitudes towards the
learning and use of the languages in specific contexts. Learners’ apparent difficul-
ties in learning certain languages may well be due to their deliberate rejection of
the socio-cultural values the languages represent.
Nevertheless, the social science perspective on applied linguistics has had disap-
pointingly little impact on the broad field that it aspires to be part of. An admira-
ble exception is perhaps Critical Discourse Analysis, where some of the language
it uses has been taken up by generic social scientists who are interested in social
change. One could say, of course, that much social science research is critical anal-
yses of public discourses. But the question remains as to how applied linguists can
have their work taken more seriously at both theoretical and practical levels by
social scientists generally, rather than having it seen as providing empirical sup-
port for existing models or policies on a limited scale.
There are a number of possible ways forward: one is to come up with some alter-
native theories of society and social structures. This would, however, risk endanger-
ing the future of applied linguistics as a discipline, as language does not have to
feature at all in theories of society and social structures. Applied linguistics would
then submerge under another branch of social science. Alternatively one could
argue that language is integral, even essential, to everything in society. Language
is certainly a critical part of the social life of individual members of society. There
is no doubt that applied linguists can make a major contribution to social science
by focusing on how members of society use the social resources, which include
languages, that are available to them to create, maintain and change social rela-
tionships and social structures.
Perhaps the most important way forward for applied linguistics is the develop-
ment of a multi-strategy approach in applied linguistic research methods (see also,
Sealey and Carter, 2004). We are not talking here about the combination of what
might be called qualitative and quantitative methods, or the macro and the micro-
analysis. We need an approach that goes beyond the traditional dichotomies, i.e.
critical of existing theories and models including existing theories of language and
of society; and that sees the social world as stratified and social phenomena as
related to one another in complex ways.
Introduction 7
3 The Present Volume: Individuals Looking to the Future
This volume is then intended to show the importance of the contribution that
applied linguistics can make to real-world problems. It does not start from some
standard model of language or language use. It does not even pretend to speak in
one unified voice. The contributions do not resemble the genre of book currently
called handbooks, which mostly have state-of-the art surveys of the field or histo-
ries of past achievements. Rather the contributors here are individuals laying out
their ideas of a future for applied linguistics in its fine tradition as a broad church
bringing together people with a primary interest in language.
The volume starts with four chapters on the relationships between languages
and broad social issues – globalization, economy, poverty and religion. Piller and
Pavlenko discuss the intersections between multilingualism and gender around
the two key domains where gender is produced and reproduced, namely economic
production and social reproduction, and how the two domains, as well as the inter-
sections, have changed in the context of globalization. Coulmas outlines what he
calls ‘econo-linguistics’ where the common concerns of economists and linguists
are dealt with systematically by focusing on the pervasive nature of language in
all human affairs. Romaine further explores the interface between linguistics and
development economics by analysing the global distribution of linguistic diversity
and poverty and the complex relations between the two dimensions of human
development. Particular attention is paid to indigenous peoples, who speak around
60 per cent of the world’s languages but are among the world’s poorest. Spolsky in
his chapter considers the interaction of language and religion from a management
policy perspective. He asks how religious institutions and leaders attempt to exer-
cise authority to modify the language practices and beliefs about language of their
followers and of others; what success might religious language managers expect to
have; and what support can family language managers expect from religion and
religious organizations.
Chapter 5 by Enfield revisits the age-old theme of the relationship between lan-
guage and culture. Instead of asking how language and culture are related, he
asks: what is the common stuff of which language and culture are built? Enfield’s
answer is that they are both cognitive and practical resources for meaningful action
on, and through, social relationships. Thus he proposes a relationship approach
to talk-in-interaction.
The next three chapters are concerned with various kinds of discourse. Gu’s
chapter offers an original conceptualization of multi-modality communication
by distinguishing land-borne situated discourse – a study of life activities, written
word-borne discourse – a study of how knowledge has been created, reserved,
revised, rejected or forgotten, air-borne and web-borne situated discourses, which
provide bridges for the land-borne and the written word-borne. Gunnarsson exam-
ines discourse in organizations and workplaces by discussing the discourse-related
problems facing managers and employees in the globalized economy. She analyses
how an ‘organizational self’ is constructed by means of discourse, and the role of

discourse for internal management and external marketing. Schäffner examines
the interface between political discourse analysis and translation studies. She seeks
to address the (in)visibility of translation in political communication, textual pro-
files of translations (as compared to their source texts), and the role of the socio-
political contexts and conditions in which translations are produced.
Gibbons in Chapter 9 outlines the central issues in the fast developing field
of Forensic Linguistics, including legal language, language legislation and linguis-
tic evidence. As he points out, Forensic Linguistics is a good example of applied
linguistics for the real world. The issues discussed in this chapter are of major sig-
nificance to those involved, whether it is people who cannot understand the legis-
lation impacting on their lives, witnesses whose testimony is distorted by linguistic
pressure tactics, minorities whose language cannot be used or who are subjected
to group vilification, or the guilty or innocent convicted by language evidence.
Gibbons shows how applied linguists can offer their professional insights into all
these issues.
The two chapters that follow demonstrate the range of applied research lin-
guists have carried out over the years. Lorch’s chapter synthesises the key issues
emerging from neurolinguistic research. Specific consideration is given to the
development of our understanding of how language is organized in the brain by
refinements in both the way research questions are framed and the manner in
which answers are sought. She further suggests a new avenue of research by what
she calls the Applied History approach, i.e. to explore current formulation of theo-
retical issues, methodological strategies and forms of argumentation in neuro-
linguistics through exemplars from the historical literature of the field. Ball and
Müller in their chapter critically examine the relationship between Clinical
Phonology, a sub-field of Clinical Linguistics, and theoretical phonology. They
argue that the current dominant approaches to phonological theory are overtly
not attempts to model the act of speech production but rather attempts to provide
elegant descriptions of data. This appears to be sometimes misunderstood by those
applying these models to disordered speech. They further argue that there is a real
need to move towards a phonological description that overtly does attempt to
model speech production, and that this provides a better principled direction for
remediation by clinicians.
The final chapter by Morgan and Woll discusses the ways in which the study of
how deaf people communicate in signs, as well as the linguistic and cognitive proc-
esses inherent in signed languages themselves, can inform other areas of applied
linguistics such as bilingualism and language acquisition. They also examine the
wider social context in which sign languages are used, including Education and
second language learning.
These thumbnail sketches do no justice to the richness and scope of the
contributions. These chapters stimulate because of the strength of their individual
views of real-world issues from an applied linguistic perspective. These are very
much individuals looking to build the future of applied linguistics grounded in its
focus on language but with a much broader concern for the real world.
Introduction 9
References
Block, D. (2009), ‘Identity in applied linguistics: The need for conceptual exploration’,
in Vivian Cook and Li Wei (eds), Contemporary Applied Linguistics. Vol. 1. Applied
Linguistics and Language Teaching in the 21st Century. London: Continuum.
Brumfit, C. J. (1995), ‘Teacher professionalism and research’, in G. Cook and
B. Seidlhofer (eds), Principle and Practice in Applied Linguistics. Oxford: Oxford
University Press.
Brumfit, C. J. (1997), ‘How applied linguistics is the same as any other science’, Inter-
national Journal of Applied Linguistics. 7, (1), 86–94.
Dornyei, Z. (2007), Research Methods in Applied Linguistics. Oxford: Oxford University
Press.
Rampton, B. (1997), ‘Returning in applied linguistics’, International Journal of Applied
Linguistics. 7, (1), 3–25.
Rampton, B. (2000), ‘Continuity and change in views of society in applied linguistics’,
in H. Trappes-Lomax (ed.), Change and Continuity in Applied Linguistics. Clevedon:
Multilingual Matters, pp. 97–114.
Sealey, A. and Carter, B. (2004), Applied Linguistics as Social Science. London:
Continuum.
CHAPTER
1
Globalization, Multilingualism, and
Gender: Looking into the Future1
Ingrid Piller and Aneta Pavlenko
1.1 Introduction
The purpose of this chapter is to discuss promising directions for future work on
multilingualism and gender in the context of globalization. Consequently, rather
than offer a comprehensive review of past work, a task we have undertaken else-
where (Pavlenko and Piller, 2001, 2007; Piller and Pavlenko, 2004), we will focus
only on issues that arise with the advent of globalization and highlight areas where
research needs to take place in the near future. We begin with a brief clarification
of how we use the terms ‘multilingualism’, ‘globalization’ and ‘gender’, and then
move on to a detailed examination of the interrelationship between language and
gender ideologies, multilingual practices and the political economy of language
and identity.
1.2 Globalization, Multilingualism, and Gender
In conceptualizing ‘multilingualism’, we follow Heller’s (2007a) view of ‘language

as a set of resources which circulate in unequal ways in social networks and discur-
sive spaces, and whose meaning and value are socially constructed within the con-
straints of social organizational processes, under specific historical conditions’ (2).
This view enables us to ask questions about how certain resources come to be val-
ued or devalued, who has access to valued resources and who is doing what with
particular resources and why.
Our understanding of the relationship between globalization and multilingual-
ism also draws on Appadurai’s (1990) description of globalization as consisting of
flows of goods, capital, communication and people. Many of these flows throw
people of widely different linguistic and cultural backgrounds into contact, either
virtual as in the case of flows of information, or face-to-face, as in the case of flows
of migration and tourism.
Applied linguists have widely debated the role of English in the globalization
process (Brut-Griffler, 2002; Pennycook, 1994, 2007; Phillipson, 1992; Sonntag,
2003). Together with sociolinguists and linguistic anthropologists, they have also
Globalization, Multilingualism and Gender 11
begun examining other topics linked to globalization, such as the commodification

of language in the new economy (Heller, 2006, 2007b), reordering of local reper-
toires due to the influx of transnational semiotic resources (Blommaert, 2005),
resistance to linguistic imperialism among subaltern groups (Canagarajah, 2005),
transnational migration (Moyer and Martin-Rojo, 2007), the interaction between
transcultural flows and language education (Block and Cameron, 2002; Kumara-
vadivelu, 2008) and ways in which struggles over language stand for struggles over
the position of nation-states and minorities in the new globalized world order
(Duchêne and Heller, 2007; May, 2001; Pujolar, 2007). These inquiries make it
increasingly clear that the service- and information-based economy, facilitated
by the new communication technologies, has created an unprecedented demand
for multilingual skills. The increased flows of tourism and migration that result
in more linguistically diverse communities around the world have also placed new
demands on language professionals (Block and Cameron, 2002). It is not acciden-
tal then that in the context of globalization, our topic is ‘multilingualism and
gender’, rather than ‘language and gender’.
Now, what does gender have to do with globalization and multilingualism? Our
understanding of the interrelationship between the three phenomena draws on
the work of the French sociologist Pierre Bourdieu (1991) in considering language
as a form of symbolic capital that can be transformed into economic and social
capital in the global marketplace. Gender, as a system of economic, social and
cultural relations, structures access to this symbolic capital, so that ‘doing multi-
lingualism’ may in itself become a gendered practice.
The way in which the ability to transform linguistic capital into economic capital
may be gendered is clearly demonstrated in a study by François Grin (2001), an
economist studying the economic value of English in Switzerland. In a quantitative
study of a sample of around 2,000 Swiss residents, this researcher found that men
who are highly proficient in English reported an average monthly income of CHF
7,636. By contrast, the average monthly income of women who are highly profi-
cient in English was CHF 4,096. Apart from the shocking difference in net earn-
ings, there was another gender difference: men’s earnings rose progressively from
‘no’ proficiency in English, via ‘low’ and ‘basic’ to ‘fluent’, i.e. the most proficient
speakers of English were rewarded the most. By contrast, for women there was a
progressive rise only from ‘no’ via ‘low’ to ‘basic’, but there was no further incre-
mental gain associated with ‘fluent’ proficiency. According to Grin (2001), this
non-linearity is due to the fact that proficient female speakers are more likely
to work part-time than their male peers. However, the effect remains even when
full-time equivalent earnings are calculated. Clearly, Swiss women who are highly
proficient in English are less successful in converting their linguistic capital into
economic capital than their male peers.
Like many other feminists, we regard work and family – that is economic pro-
duction and social reproduction – as the key areas where gender is produced and
reproduced (Grob and Rothmann, 2005). Consequently, we will structure our
examination around gender in multilingual employment and domestic sites – sites
that crucially intersect and overlap, and sites that have been subject to enormous
12 Ingrid Piller and Aneta Pavlenko
changes in recent years, as part of developments that can broadly be termed as

globalization.
1.3 Gender and Multilingualism in Employment
To better understand the relationship between gender, multilingualism and

employment it is useful to distinguish between different types of work. A crucial
difference is of course the one that holds between paid and unpaid work – a differ-
ence that has long been a gender default line, with paid work being more easily
accessible to men, and unpaid work – such as subsistence farming, household
work, child-rearing and other forms of care work – being women’s work. We will
discuss the ways in which multilingualism can end up as unpaid women’s work
later in the chapter.
At present, in most societies around the world, paid work is no longer the exclu-
sive domain of men, and women are as likely as men to take up paid employment.
However, crucial differences remain in the type of work that is being taken up.
Somewhat crudely, we find it useful to distinguish between three different types of
employment because they can be associated with distinct patterns of gendered
multilingualism. These are (1) employment in agriculture, primary extraction and
manufacturing; (2) employment in the ‘new’ information- and service-based eco-
nomy and (3) paid reproductive work. We will discuss the ways in which multi-
lingualism and gender intersect in each of these arrangements in turn, pointing
to productive directions for future work.
1.3.1 Gendered Bilingualism in the ‘Old’ Economic Order
At the base of the so-called ‘old’ economic order is work in agriculture, primary
extraction and production. In patriarchal societies paid work in these areas is
taken up by men, while reproductive work is assigned to women. In the 1970s and
1980s traditional minority communities, in transition between subsistence agricul-
ture and manufacturing work, became the first locus of attention of pioneering
scholars in the new field of multilingualism and gender such as Susan Gal (1978),
Jane Hill (1987) and Brigitte Schlieben-Lange (1977). They found that within
such traditional arrangements men encounter more opportunities to become
bilingual in the majority language, since they have more chances for interactions
in the majority language outside the subsistence farm and/or home. Women, on
the other hand, have fewer opportunities for such interaction and also more
limited access to education. As a result, indigenous women in Latin America, for
instance, were less likely to be proficient in the majority language, Spanish, than
indigenous men (e.g. Hill, 1987; Spedding, 1994). For instance, Hill (1987) studied
the use of Spanish and Mexicano, or Nahuatl, in rural communities in the region
of the Malinche Volcano in Mexico. Women in this community had less access to
education than men and spoke less Spanish. As a result, it was difficult for them
to join the paid labour force, for which the use of Spanish was crucial, and they
had limited opportunities to practice whatever Spanish they knew. In such cases,
a vicious circle may emerge in which limited proficiency in the majority language
limits access to (non-domestic) paid employment, which, in turn, limits interac-
tional opportunities in the majority language.
Minority women often relate their bleak social and economic situation to their
use of the minority language. Consequently, they may be ready to spearhead lan-
guage shift to the majority language if they have the opportunity. This pattern was
first demonstrated by Gal (1978) in her work on the bilingual town of Oberwart in
Austria. In the minority Hungarian community there, young women led the shift
towards German. They were motivated by a symbolic link between German and
industrial work that was becoming available at the time. For these peasant women,
German-speaking factory work represented a significant improvement over the
drudgery of Hungarian-speaking peasant life. McDonald (1994) reports similar
findings for Breton peasant women who at the time of the study were moving away
from rural Brittany in search for a better life in urban centres. For these women,
language shift from Breton to French represented a symbolic journey ‘from cow-
shit to finery’ (91).
It appears then that in the context of the ‘old’ economy the link between
employment, gender and bilingualism is indirect: the job opportunities available
to minority communities are agricultural and working class jobs, and minority
women find themselves in a doubly marginalized position – they are disadvantaged
as minority members, and as women. In such circumstances, minority women may
‘misrecognize’ – in Bourdieu’s (1991) term – language as the cause of their limited
prospects, and consider language shift as a way to improve their employment
opportunities, and their economic and social prospects more generally.
When indigenous languages are oral, the demand for the majority language –
or a lingua franca – and the accompanying literacy skills becomes even more
urgent. This situation is illustrated by Egbo (2004) with an example from Nigeria
where English functions as a lingua franca and the language of education and
literacy for members of approximately 250 ethnic groups who speak as many as
400 languages and dialects. In a study of two rural communities in south-western
Nigeria, the researcher found that all but one of the literate women were engaged
in the formal wage sector, and as a result had higher incomes and more economic
independence than their non-literate peers engaged in labour-intensive subsist-
ence farming. There was also a remarkable difference between the living condi-
tions of the two groups of women: the literate women worked fewer hours than the
non-literate ones and had enough time for leisure, which in itself is an indicator of
an enhanced quality of life.
As the key language of globalization, English may be viewed as an important
form of symbolic capital even in contexts where it does not function as a majority
language or a lingua franca. A study by Niño-Murcia (2003), for instance, suggests
that in Peruvian rural communities English is prized even higher than Spanish and
imagined as the road to a better life. Rural Peruvians are aware that Peruvian cities
are not offering them better chances because they will be seen as inferior, racial-
ized Quechua speakers. The same awareness does not extend to what it means to
be a Latino immigrant in the United States. Consequently, learning English takes

on a greater significance than learning Spanish because English is imagined as
a means of escape from the harsh conditions of life in rural Peru.
Women in communities whose livelihoods depend on agriculture, primary
extraction and manufacturing are facing ever greater burdens as they increasingly
have to enter the paid workforce ‘without relinquishing any of their responsibilities
in taking care of the family, fetching (sometimes buying) water, cooking and clean-
ing’ (Seabrook, 2007: 104). The many ways in which language and literacy open or
close opportunities in such communities in the grip of economic forces beyond
their control is an important area of future research. Women in the Global South
grow 60–80 per cent of the food but own less than 2 per cent of the land (Food and
Agricultural Organization of the United Nations, 2006; quoted in Patel, 2007: 302).
What are the implications of the (in)ability to read and write (in languages of wider
communication) for the livelihoods of these women and their families?
A related area where more research is needed relates to the migrations from
rural communities to the cities in search of a better life: ‘every year, 70 million
people are added to the urban population of developing countries. World-wide,
however, it is estimated that about 200 million people leave their home in search
of a livelihood – 3 per cent of the population of the earth’ (Seabrook, 2007: 44).
Again, (lack of) access to linguistic resources is one aspect that may help or hinder
the search for new livelihoods.
Finally, we would argue that there is an urgent need for such research particu-
larly in communities of the Global South. The accusation that people in the Global
North are by and large unaware and unconcerned about the people in the Global
South who produce a good part of their food, mine their primary materials and
manufacture a large part of their products (Patel, 2007; Seabrook, 2007), holds
to a degree also for linguistic and educational research. Despite the fact that some
of the early ground-breaking research into multilingualism and gender in employ-
ment was conducted in Latin America (e.g. Hill, 1987; Spedding, 1994), there is
only a very limited continuation of this research in the context of accelerated eco-
nomic globalization, particularly as far as communities in Africa and Asia are con-
cerned (see however Egbo, 2004; Williams, 2006).
1.3.2 Gendered Language Work in the “New” Economic Order
Under the ‘old’ economic order, the road to economic advancement often included
the journey from monolingualism in the minority language, to bilingualism, to
monolingualism in the majority language. In the ‘new’ economic order, where the
service industry and tourism occupy central places, the road to economic advance-
ment often involves the journey to multilingualism in the minority language, the
majority language, English, and any number of other languages that might have
cache in a given market.
The tourist flows of the globalized world enabled some rural minority commu-
nities and particularly women in these communities to turn their languages into
a commodity: the commodity being, ironically, a marker of authenticity in tourism

contexts. A good example of this process is provided by Nushu (or Nü Shu), the
women’s script used by women in China’s Hunan province. Often used as a text-
book example of the only ‘women’s language’ in the world, this script was tradi-
tionally used as a secret language by ‘sworn sisters’ to lament the hardships of
having to leave one’s birth family at marriage. As the lot of women improved with
the Chinese Revolution, including access to literacy in the standard Chinese script,
Nushu fell into disuse. However, today there is a significant tourist market for
Nushu, with a language school open only to women in Jianyong City. Jianyong has
emerged as an important destination for cultural tourism, and souvenirs include
Nushu embroidery and other language-related artefacts: ‘tourists and academics
come to hear the women sing, sew and write’ (Watts, 2005).
In the service industry, multilingual skills are even more obviously commodi-
fied – they become language work, i.e. any form of labour where the work that is
remunerated is linguistic (e.g. language teachers, translators and interpreters, call
centre staff). The language workforce is heavily feminized. About two-thirds
of workers in European and North American call centres are young women,
recruited ‘on account of presumed “feminine” characteristics, primarily aptitude
for communication and personal skills’ (Glucksmann, 2004: 797). The evidence
from other sites of multilingual language work, be it English teachers in Germany
(Gohrisch, 1998) or call centre workers in Canada (Heller, 2005, 2007b; Roy,
2003), points in the same direction: it is a heavily feminized industry. This is not
surprising: as Cameron (2000a,b, 2005) has repeatedly shown in her work in mono-
lingual English call-centres, communication styles that are valued in these jobs are
those that have traditionally been considered ‘female speech styles’.
However, feminization of multilingual language work is not a cause for celebra-
tion: most language workers, like most service workers generally, work in casual,
part-time, junior and relatively poorly paid positions with little job security (Massey,
2004). In fact, as Heller (2003) demonstrates with regard to bilingual service work-
ers in Canada, their prospects for advancement may be more limited than those of
their monolingual peers because the organization of workplaces is such that it
ensures that bilingual staff will remain in positions where their multilingual skills
are most useful, namely in ‘front-line’ service work. Language work also fits in well
with typical female biographies: its casual nature makes it possible to combine
work with ‘family’, i.e. unpaid reproductive work.
In sum, recent decades have been characterized by fundamental changes in the
economic order. These changes have heavily impacted on gender and multilin-
gualism, both independently and in interaction. When minority languages were
tied to authenticity, minority women often ended up less multilingual than their
male peers, their monolingualism making them in effect ‘living symbols of tradition’
(Cameron, 1992: 202). In the ‘new’ economic order ‘authenticity’ now comes with
a price tag, which means that minority communities, and particularly minority
women, have sometimes been able to find economic value in putting their minor-
ity language on sale (Watts, 2005). New job opportunities calling for multilingual
skills have emerged in the service industry – these positions are disproportionately
occupied by women and are also casual, part-time, junior, relatively poorly paid
and with little job security.
So far, the study of commodification of minority languages as signifiers of
authenticity in cultural and heritage tourist destinations – as well as substantial
participation of women in these projects –have received only limited sociolinguis-
tic attention (but see Heller (2005) for a notable exception). Future studies need
to engage critically with the commodification of multilingual skills. There is a need
for more industry-specific ethnographic work considering language work in the
service and knowledge economy. Additionally, more quantitative macro-economic
investigations in specific national economies, along the lines of Grin (2001), are
also needed.
1.3.3 Paid Reproductive Work
Whatever the shortcomings of language work may be, there is another feminized
job sector that is significantly more exploitative, and that is paid reproductive work,
such as child care, elder care and domestic work. Ehrenreich and Hochschild’s
(2002a) collection entitled Global Woman: Nannies, Maids, and Sex Workers in the
New Economy offers illuminating insights into the profound impact globalization
has had on gender relations worldwide. As a consequence of the structural adjust-
ment measures imposed by the International Monetary Fund, the World Bank and
related bodies, the gap between the rich and the poor has further widened on the
globe. This ever-increasing inequality is well documented (e.g. Munck, 2005; Patel,
2007; Seabrook, 2007) despite the rhetoric that often heralds globalization as
some form of development aid. At the same time that the economic pressures
on families in the global South increase, the global media bring images of consum-
erism to almost every household in the world, in a kind a ‘material striptease’
(Ehrenreich and Hochschild, 2002b: 8).
One of the consequences of neoliberal economic regimes in conjunction with
the iconization of consumerism is that women, more than ever before, are on the
move. In the late 1990s, the number of migrant women equalled the number of
migrant men for the first time in recorded history (Ehrenreich and Hochschild,
2002a). Female migratory flows mainly originate in South East Asia, Latin America
and Africa, and are directed towards Europe, the Gulf States and North America.
The overwhelming majority of these ‘new’ female economic migrants do typical
‘women’s work’, i.e. reproductive work such as domestic work, child care and elder
care, as well as sex work.
The booming mail-order bride industry can be considered part of the same
phenomenon (Piller, 2007). These women are linked in what has come to be
described as ‘global care chains’ – where, for instance, an affluent North American
dual-income couple hires a live-in nanny from the Philippines to care for their
children. This woman from a middle-class background has to leave her own chil-
dren behind in the Philippines, and hires another woman, an internal migrant
from a rural area in the Philippines to care for them. This woman in turn has left
her own care responsibilities behind, typically as unpaid work for her mother,
mother-in-law, sister or another female relative. The need for reproductive workers
has re-emerged in the global North not because women there have taken up paid
work outside the home but rather because men there have NOT taken up unpaid
work in the home (Blair-Loy and Jacobs, 2003). For many women in the Global
North migrant reproductive workers provide them with a way out of ‘the second
shift’ (Hochschild, 2003). What used to be a gender divide – domestic work – is being
replaced by a class and race divide that is also gendered (Grob and Rothmann,
2005).
How does language play out in these global domestic scenarios? On the basis
of sociological inquiry, (e.g. Anderson, 2000; Chang, 2000; Hondagneu-Sotelo,
2001; Parrenas, 2001, 2005), it is clear that the women who join the ‘care drain’ are
mostly unlikely to speak the language of the host country or they may have only
limited proficiency in it. As Anderson (1997: 42) puts it: ‘domestic work does not
require the worker to speak the language of the host country.’ Indeed, from the
employer’s perspective, non-existent or limited proficiency in the majority lan-
guage may be an advantage: reproductive work involves access to the most intimate
details of family life, and limited proficiency may serve as a way to keep up the pre-
tence of distance. Limited proficiency in the majority language may also be one
way for the employer to regard the reproductive worker as inferior and to rational-
ize their unequal status, as one Filipina interviewee in a study of domestic workers
in Toronto put it: ‘they think you are as stupid as your English is’ (England and
Stiell, 1997). Clearly, limited proficiency is a way to curtail the limited freedoms of
migrant domestic workers even further. In a recent Australian court trial migrant
sex workers – a related sector – were found to have been held in slavery because
‘while the women were not kept under lock and key, they could not run away as
they had no money, no passport, limited English and were told to avoid immigra-
tion authorities’ (AAP, 2006). The quote exemplifies how crippling limited profi-
ciency can be, particularly for reproductive workers, as it often co-occurs with other
constraints upon their freedom such as illegal visa status or fear of deportation.
Limited or non-existent proficiency in the majority language exacerbates the
vulnerability of migrant reproductive workers. It is also implicated in making
reproductive work more strenuous. We are not aware of a systematic exploration
of multilingualism in labour that is intrinsically emotionally strenuous, as well as
physically taxing. However, it is clear that migrant reproductive workers are
extremely vulnerable to emotional stress: they often leave their own children
behind to care for the children of others, and the emotional costs of transnational
motherhood are well documented (Dreby, 2006). In addition to guilt as a key emo-
tion with regard to the children left behind, care work, by its very nature, involves
a high degree of emotional work. Not being able to communicate with the persons
one cares for may add to the emotional difficulty of the work. In the literature we
consulted, lack of linguistic proficiency is a re-occurring aspect of interviews with
migrant care workers when they recount the hardships of their work. For instance,
Veronica, a 24-year-old dentistry student from Ecuador, who, as an undocumented
immigrant, works as a live-in caretaker for an elderly man in Tel Aviv, states:
I started right away working as a live-in caretaker of an old man. The first week
was a nightmare. For someone like me who had never done this type of job and,
on top of it, in a completely unknown country and without knowing a single
word in Hebrew!! . . . I was supposed to be at his side round the clock and stay
at home. I could not help but cry the whole week. (Raijman et al., 2003: 738)
In some cases, the language proficiency of reproductive workers may be one of

their assets. Nannies from the Philippines are preferred over those from other
South-East Asian countries in Continental Europe and the Gulf States because
they can speak to their charges in English. The advice literature on raising bilin-
gual children (see below) often contains the suggestion to hire a nanny who is a
native speaker of the minority language in order to increase the child’s exposure
to that language. A particularly interesting example of the valorization of the
minority language in domestic workers can be found in Kazakhstan, a former
Soviet republic and now a nation-state. During the Soviet period, urban Kazakh
elites became heavily russified, while the rural populations continued to be mono-
lingual in Kazakh. In independent Kazakhstan, Kazakh has acquired new value
and prestige. Consequently, migrants arriving in the cities from rural areas in
the south and taking up reproductive work are bringing with them valuable lin-
guistic capital: interactions with these women contribute to the increase in levels
of Kazakh-language competence among their predominantly Russian-speaking
Kazakh employers (Smagulova, 2008).
However, independent of whether their existing linguistic proficiencies are
valued or not, migrant reproductive workers tend to find themselves in the same
depressing and isolated situation that full-time unpaid homemakers find them-
selves in – a staple of the feminist critique of gender arrangements since Friedan
(1963). While the isolation experienced by migrant reproductive workers may be
similar to those of the American housewife of the 1950s, illegal status, lack of lin-
guistic proficiency in the majority language, and related aspects clearly exacerbate
their dependency and vulnerability. Consequently, migrant reproductive workers
sometimes regard living-out arrangements as their only ‘career’ option: not only
does it give them greater control over the hours they work, and allows them to
separate work and private life, it also crucially allows them to learn the language of
the host society.
In recent decades, the role of gender in language and literacy learning and
use by immigrants in Western countries has become a topic of lively research. One
argument that emerged in this area recently is that as time goes by women may
overtake men in the language learning arena (e.g. Baynham, 2005; Gordon, 2004;
Vitanova, 2004) and that interactions linked to child care and other domestic work
may open more language learning opportunities for them than wage work side-by-
side with other immigrants (e.g. Goldstein, 2001; Gordon, 2004; Norton, 2000).
What is not clear yet is whether these gains can be systematically translated into
social advancement (see, however, Vitanova, 2004, for a study of language learning
and work trajectories of four immigrant couples). The study of language learning
and use by international reproductive workers offers great promise in this debate.
The field also needs longitudinal and socio-economically oriented studies to exam-
ine whether and how the mastery of the majority language is converted into eco-
nomic advantages by immigrant men and women and how the conversion process
is mediated not only by gender but also by class and race.
1.4 Romancing Multilingualism
In addition to work, the family created on the basis of heterosexual coupling is

the second key site where gender is produced and reproduced. Paid and unpaid
reproductive work are linked in numerous ways, as expressed, for instance in the
global care chain, described above. We will now turn away from paid employment
to focus on the implication of multilingualism in social reproduction. We will
address two aspects where gender and multilingualism intersect in the family,
namely heterosexual romance and childrearing.
1.4.1 Language Desire
Migration flows in search of employment are only one kind of transnational flows.
Another flow directly relevant for our discussion is language tourism that most
commonly takes place in the form of study abroad, organized or individual. One
of the many motivations that may drive language sojourns abroad is language desire,
a desire to become a legitimate speaker of a particular language, to construct
a new social and gender identity in this language and/or to form a romantic rela-
tionship with a speaker of the desired language (possibly to get the coveted legiti-
macy). The term originates in Piller’s (2002) and Piller and Takahashi’s (2006)
work; in the latter it is also linked to the notion of power. Following Foucault
(1977, 1980), Piller and Takahashi (2006) argue that the workings of power include
the inculcation of desires that drive individuals to modify their bodies, personali-
ties, life trajectories, or, for that matter, linguistic repertoires. As a result, language
desire, too, may be hegemonic and an instrument through which individuals con-
spire in their own oppression.
The first example of ways in which public discourses structure private desires
comes from research on Americans on study abroad. In the United States, study
abroad has traditionally been perceived as a gendered enterprise, which sends
‘wealthy women to academically weak European programs established in a frivo-
lous Grand Tour tradition’ (Gore, 2005: 24). The gendered trend continues to this
day: according to Kinginger (2008), in 2005–2006, women constituted 65.5 per cent
of the participants in organized study abroad. A particularly popular location for
these students is France, whose irresistible eternal femininity has been commodi-
fied in numerous publications for American women on how to discover their
‘inner French girl’ (Ollivier, 2003) or eat for pleasure – like French women do! –
without getting fat (Giuliano, 2005).
Kinginger’s (2008) ethnographic study of American students in France reveals
that these dominant ideologies of gender and embodiment are internalized by
many of the study participants yet with different consequences. Men can position
themselves as appreciative consumers of stylish French women and benefit from
their interactions gaining in linguistic proficiency (for similar findings on male
American students in Russia, see Polanyi, 1995). Women who want to depart from
their Americanness and resemble the French are forced to question their own
body image and behaviours. Some adapt and are rewarded with a rich social net-
work and eventually with linguistic gains. Others develop feelings of resentment
towards ‘snotty’ French women, behaviours perceived as sexual harassment and
even advertisements portraying naked women. These ads, according to one of the
study participants, made her hate going outside. This participant ended up mak-
ing almost no contact with French speakers and made no visible linguistic gains by
the end of her semester abroad.
Importantly, the participants in Kinginger’s study are middle- and upper-middle-
class Westerners on a study-abroad program in another Western country. When
racialized non-Westerners come into play, the relations of power change dramati-
cally, as seen in an ethnographic study of Japanese women on a study-abroad expe-
rience in Australia (Piller and Takahashi, 2006; Takahashi, 2006). The participants
in the study often had spent their youth in Japan fantasizing about romantic
relationships with Hollywood stars such as Tom Cruise or Brad Pitt. Once they
followed their dreams and embarked upon an overseas experience in Sydney, they
found that Sydney was by no means the welcoming place they had imagined. There,
they had imagined themselves becoming fluent in English easily and effortlessly.
When this did not happen and they found that it was actually difficult to even enter
into interactions with English speakers, their search for an English-speaking roman-
tic partner gained new urgency. An English-speaking romantic partner seemed a
way to ensure regular interactions in English. The English speakers they sought
out as potential romantic partners were first and foremost White, so that the link
between symbolic ownership of English and race was reproduced. All of the partic-
ipants stayed on in Sydney beyond the originally intended period, which led most
of them into a financially difficult or dependent situation. As the expected rewards
of these sacrifices – fluency in English, full participation in Sydney’s fashionable
society, a handsome White boyfriend – were not forthcoming, desire could even
give way to depression.
While all of the Japanese participants in the study exerted considerable agency
in their choices of potential romantic partners, the power imbalance inherent in
the emotional relationship that sets up the West as an object of desire and Japanese
women as the desirers serves powerful market interests. This is the language school
and study overseas market and, on a broader level, the consumption of all the
branded goods and services associated with a desirable Western lifestyle. ‘Relation-
ship English’ – English for the specific purpose of conducting a romantic relation-
ship – has become the most recent addition to the offerings of the ever-expanding
TESOL industry. Future work in this area will need to explore ways in which lan-
guage desire is constituted in public and private discourses in a wider range of
contexts, asking whose interests are served – or disserved – in particular cases.
1.4.2 Multilingual Child-Rearing
Searching for a romantic partner is only one form of language desire. Another
form of language desire is bilingual child-rearing to which we will now turn. Like
all parents, bilingual parents are the target audience for the burgeoning parenting
advice genre. Advice is offered on almost any conceivable aspect of the parent–child
relationship from conception to training physical superiority and does of course
also include bilingual parenting. Two major strands of bilingual child-rearing
advice can be distinguished: there is the fashion of teaching babies sign language
before they learn how to speak (e.g. Acredolo et al., 2002; Garcia, 2002), and there
is advice on raising children bilingually in two or more oral languages (e.g. Barron-
Hauwaert, 2004; Cunningham-Andersson and Andersson, 1999; Harding-Esch and
Riley, 2003; King and Mackey, 2007; Tokuhama-Espinosa, 2003). These and many
related products (web sites, internet forums, newspaper articles etc.) regularly
extol the wonderful advantages of raising children bilingually.
We do not wish to debate the virtues of being exposed to more than one lan-
guage from an early age but we do want to discuss what ‘bilingual parenting’ means
from a gender perspective. Most of the advice literature uses admirably gender-
neutral language, and presents bilingual ‘parenting’ as outside of the realm of the
gendered division of family labour. However, in the same way that sociologists have
observed that men have not taken on their fair share of reproductive work at the
same time that women have taken on paid employment outside the home, it is not
clear whether men have taken on their share of parenting work, including lan-
guage work in bilingual parenting.
A negative answer to this question comes from Okita’s (2002) ethnographic
research with Japanese mothers in the United Kingdom. The women in the study
had – or had originally had – the ambition to raise their children bilingually in
English and Japanese. As they were all married to British men with limited or no
proficiency in Japanese, it was clear that they would have to take on the task of
teaching their children Japanese. One of the many consequences this had was that
the stay-at-home-mum model became the only viable option during their children’s
preschool years to ensure sufficient exposure. Alternative care arrangements –
from having family and friends look after the children to institutional childcare
arrangements – necessarily meant curtailing input in Japanese. Thus, bilingual
mothering was hard work for these women, and it was the harder because it was
not recognized as work – like many other forms of emotional and reproductive
work it is ‘invisible work’ – invisible even to some of the fathers in the study. Indeed,
one of the reasons why bilingual child-rearing became so stressful for these women
is because their partners failed to recognize their work and support them. This
researcher is clear about the gendered implications bilingual child-rearing has for
mothers: it ‘lead[s] to disempowerment, intensified pressure, guilt and personal
trauma’ (Okita, 2002: 230).
Okita’s (2002) conclusions can also be extended to parenting in traditional
minority and immigrant communities where minority language maintenance is
women’s responsibility and, for Cameron (1992), yet another way to keep women
within the confines of the house. In future work, it will be important to explore
how childhood bilingualism plays out in other language communites, and also in
non-traditional family structures, along the lines of Schüpbach (2005), who found
that a single mother in her study was much more successful in transmitting the
minority language – Swiss German in Australia in this case – to her daughter than
any of the other participating Swiss German migrants who lived in more tradi-
tional family arrangements.
1.5 Conclusion
Multilingualism is a form of practice, and it is a gendered practice. We have struc-

tured our discussion of the intersections between multilingualism and gender
around the two key domains where gender is produced and reproduced, namely
economic production and social reproduction. These have changed considerably
in the context of globalization, and so has the interrelationship between multilin-
gualism and gender. We described two broad ways in which gender structures
multilingualism in these domains. First, gender structures access to language as
symbolic capital, which is particularly apparent under the ‘old’ economy where
gender relationships have often been based on the model of the male breadwinner
and the stay-at-home housewife. Under the ‘new’ economy, ‘doing multilingualism’
may in itself become a (commodified) gendered practice. Language work, language
learning, and bilingual child-rearing have all become sites that are implicated in
the reproduction of hegemonic gender ideologies. Be it ‘authentic’ minority lan-
guages on display in heritage tourism, be it the marketing of English study as a way
to win the heart of an English-speaking Prince Charming or be it bilingual child-
rearing as an addition to the numerous invisible tasks of ‘women’s work’ – multi-
lingualism becomes a discursive space were hegemonic notions of femininity and
masculinity are produced and reproduced. These are domains where we see our
own profession – as language teachers and academic linguists – directly implicated:
extolling the virtues of language learning, studying abroad, or childhood bilin-
gualism as if these were intrinsically valuable without reference to their social con-
text is at best naïve, and at worst politically dangerous.
We hope to have demonstrated in this chapter that there is an urgent need to
shift our professional perspective away from language per se onto social practices –
be they gendered, raced or classed – in order to understand how language and
literacy are a key part of the work choices and personal choices available to com-
munities and individuals.
Note
1 A different version of this paper has appeared as Piller and Pavlenko (2007).
References
AAP (2006), ‘Sex Slavery: First Woman Jailed. Sydney Morning Herald, June 9, 2006,
http://www.mg.co.za/article/2005-12-07-mother-tongue.
Acredolo, L., Goodwyn, S. and Abrams, D. (2002), Baby Signs: How to Talk with Your Baby
Before Your Baby Can Talk. New York: McGraw-Hill.
Anderson, B. (1997), ‘Servants and slaves: Europe’s domestic workers’, Race and Class,
39, (1), 37–49.
Anderson, B. (2000), Doing the Dirty Work? The Global Politics of Domestic Labour. London:
Zed Books.
Appadurai, A. (1990), ‘Disjuncture and difference in the global cultural economy’,
Theory, Culture and Society, 7, 295–310.
Barron-Hauwaert, S. (2004), Language Strategies for Bilingual Families: The One-Parent
One-Language Approach, Parents & Teachers Guides. Clevedon, UK: Multilingual Matters.
Baynham, M. (2005), ‘Network and agency in the migration stories of Moroccan
women’, in Baynham, M. and A. De Fina (eds), Dislocations/relocations: Narratives of
Displacement. Manchester, UK: St. Jerome Publishing, pp. 15–35.
Blair-Loy, M. and Jacobs, J. (2003), ‘Globalization, work hours, and the care deficit
among stockbrokers’. Gender and Society, 17, (2), 230–249.
Block, D. and Cameron, D. (eds) (2002), Globalization and Language Teaching. London/
New York: Routledge.
Blommaert, J. (2005), ‘In and out of class, codes and control: Globalization, discourse,
and mobility’, in Baynham, M. and A. De Fina (eds), Dislocations/Relocations: Narra-
tives of Displacement. Manchester, UK: St. Jerome Publishing, pp. 128–143.
Bourdieu, P. (1991), Language and Symbolic Power. Cambridge: Polity.
Brut-Griffler, J. (2002), World English: A Study of its Development. Clevedon, UK: Multi-
lingual Matters.
Cameron, D. (1992), Feminism and Linguistic Theory. 2nd ed. London: Macmillan.
Cameron, D. (2000a), Good to Talk? Living and Working in a Communication Culture.
London: Sage.
Cameron, D. (2000b), Styling the Worker: Gender and the Commodification of Lan-
guage in the Globalized Service Economy. Journal of Sociolinguistics 4, 3, 323–347.
Cameron, D. (2005), ‘Communication and commodification: Global economic change
in sociolinguistic perspective’, in Erreygers, G. and G. Jacobs (eds), Language, Com-
munication and the Economy. Amsterdam: John Benjamins, pp. 9–23.
Canagarajah, S. (ed.) (2005), Reclaiming the Local in Language Policy and Practice.
Mahwah, NJ: Erlbaum.
Chang, G. (2000), Disposable Domestics: Immigrant Women Workers in the Global Economy.
Cambridge, MA: South End Press.
Cunningham-Andersson, U. and Andersson, S. (1999), Growing up with Two Languages:
A Practical Guide. London: Routledge.
Dreby, J. (2006), ‘Honor and virtue: Mexican parenting in the transnational context’,
Gender and Society, 20, (1), 32–59.
Duchêne, A. and Heller, M. (eds) (2007), Discourses of Endangerment: Ideology and Interest
in the Defence of Languages. London/New York: Continuum.
Egbo, B. (2004), ‘Intersections of literacy and construction of social identities’, in
Pavlenko, A. and Blackledge A. (eds), Negotiation of Identities in Multilingual Contexts.
Clevedon, UK: Multilingual Matters, pp. 243–265.
Ehrenreich, B. and Hochschild, A. (eds) (2002a), Global Woman: Nannies, Maids, and
Sex Workers in the New Economy. New York: Metropolitan Press.
Ehrenreich, B. and Hochschild, A. (2002b), ‘Introduction’, in Ehrenreich, B. and
A. Hochschild (eds), Global Woman: Nannies, Maids and Sex Workers in the New
Economy. London: Metropolitan Press, pp. 1–13.
England, K. and Stiell, B. (1997), ‘“They think you’re as stupid as your English Is”:
Constructing foreign domestic workers in Toronto. (Citizenship and International
Migration)’. Environment and Planning, A29, (2), 195–215.
Foucault, M. (1977), Discipline and Punish: The Birth of the Prison. New York: Pantheon.
Foucault, M. (1980), Power/Knowledge: Selected Intereviews and Other Writings, 1972–1977.
New York: Pantheon.
Friedan, B. (1963), The Feminine Mystique. New York: W. W. Norton.
Gal, S. (1978), ‘Peasant men don’t get wives: Language and sex roles in a bilingual
community’, Language in Society, 7, (1), 1–17.
Garcia, J. (2002), Sign with your Baby: How to Communicate with Infants before They
can Speak. Mukilteo, WA: Northlight Communications.
Guiliano, M. (2005), French Women Don’t Get Fat: The Secret of Eating for Pleasure.
New York: Alfred Knopf.
Glucksmann, M. A. (2004), ‘Call configurations: Varieties of call centre and divisions
of labour’. Work, Employment and Society, 18, (4), 795–811.
Gohrisch, J. (1998), ‘Zur Situation des Anglistischen Nachwuchses [the Situation of
Junior Academics in English Studies]’, Anglistik: Mitteilungen des Verbandes Deutscher
Anglisten, 9, (2), 131–140.
Goldstein, T. (2001), ‘Researching women’s language practices in multilingual work-
places’, in Pavlenko, A., Blackledge, A., Piller, I. & M. Teutsch-Dwyer (eds), Multilin-
gualism, Second Language Learning, and Gender. Berlin/New York: Mouton De Gruyter,
pp. 77–101.
Gordon, D. (2004), ‘“I’m tired. You clean and cook.” Shifting gender identities in
second language socialization’, TESOL Quarterly, 38, (3), 437–457.
Gore, J. E. (2005), Dominant Beliefs and Alternative Voices: Discourse, Belief, and Gender in
American Study Abroad. New York: Routledge.
Grin, F. (2001), ‘English as economic value: Facts and fallacies’. World Englishes, 20, (1),
65–78.
Grob, R. and Rothmann, B. (2005), ‘Parenting and inequality’, in Romero, M. and
E. Margolis (eds), The Blackwell Companion to Social Inequalities. Oxford: Blackwell,
pp. 238–256.
Harding-Esch, E. and Ph. Riley (2003), The Bilingual Family: A Handbook for Parents. 2nd
ed. Cambridge: Cambridge University Press.
Heller, M. (2003), ‘Globalization, the new economy, and the commodification of lan-
guage and identity’, Journal of Sociolinguistics, 7, (4), 473–492.
Heller, M. (2005), ‘Language, skill and authenticity in the globalized new economy’,
Noves SL: Revista de Sociolingüística, 4.
Heller, M. (2006), Linguistic Minorities and Modernity. Second ed. London/New York:
Continuum.
Heller, M. (2007a), ‘Bilingualism as ideology and practice’, in Heller, M. (ed.), Bilin-
gualism: A Social Approach. Hampshire/New York: Palgrave, pp. 1–22.
Heller, M. (2007b), ‘Gender and bilingualism in the new economy’, in McElhinny, B. (ed.),
Words, Worlds, and Material Girls. Berlin/New York: Mouton De Gruyter, pp. 287–304.
Hill, J. (1987), ‘Women’s speech in modern Mexicano’, in Philips, S., Steele, S. and
Ch Tanz (eds), Language, Gender, and Sex in Comparative Perspective. Cambridge:
Cambridge University Press, pp. 121–160.
Hochschild, A. (2003), The Second Shift. New York: Penguin.
Hondagneu-Sotelo, P. (2001), Doméstica: Immigrant Workers Cleaning and Caring in the
Shadows of Affluence. Berkeley: University of California Press.
King, K. and Mackey, A. (2007), The Bilingual Edge: Why, When, and How to Teach Your
Child a Second Language. New York: Collins.
Kinginger, C. (2008), Language Learning in Study Abroad: Case Histories of Americans in
France. Blackwell: Modern Language Journal Monograph Series.
Kumaravadivelu, B. (2008), Cultural Globalization and Language Education. New Haven,
CT: Yale University Press.
Massey, D. (2004), ‘Space, place and gender’, in Miles, M., Hall, T. and I. Borden (eds),
The City Cultures Reader. London and New York: Routledge, pp. 307–310.
May, S. (2001), Language and Minority Rights: Ethnicity, Nationalism, and the Politics of
Language. Harlow, UK: Longman.
McDonald, M. (1994), ‘Women and linguistic innovation in Brittany’, in Burton,
P., Dyson, K. and Sh. Ardener (eds), Bilingual Women: Anthropological Approaches to
Second-Language Use. Oxford and Providence, RI: Berg, pp. 85–110.
Moyer, M. and L. Martin-Rojo (2007), ‘Language, migration, and citizenship: New chal-
lenges in the regulation of bilingualism’, in Heller, M. (ed.), Bilingualism: A social
Approach. Hampshire/New York: Palgrave, pp. 137–160.
Munck, R. (2005), ‘Social exclusion: New inequality paradigm for the era of globali-
zation?’ In Romero, M. and E. Margolis (eds) The Blackwell Companion to Social
Inequalities. Oxford: Blackwell, pp. 31–49.
Niño-Murcia, M. (2003), ‘“English is like the dollar”: Hard currency ideology and the
status of English in Peru’, World Englishes, 22, (2), 121–141.
Norton, B. (2000), Identity and Language Learning: Gender, Ethnicity, and Educational
Change. Harlow, UK: Longman.
Okita, T. (2002), Invisible Work: Bilingualism, Language Choice and Childrearing in Inter-
married Families. Amsterdam: Benjamins.
Ollivier, D. (2003), Entre Nous: A Woman’s Guide to Finding Her Inner French Girl.
New York: St. Martin’s Griffin.
Parrenas, R. (2001), Servants of Globalization: Women, Migration and Domestic Work.
Stanford, CA: Stanford University Press.
Parrenas, R. (2005), Children of Global Migration: Transnational Families and Gendered
Woes. Stanford, CA: Stanford University Press.
Patel, R. (2007), Stuffed and Starved: Markets, Power and the Hidden Battle for the World Food
System. London, Portobello Books.
Pavlenko, A. and Piller, I. (2001), ‘New directions in the study of multilingualism,
second language learning, and gender’, in Pavlenko, A., Blackledge, A., Piller, I.
and M. Teutsch-Dwyer (eds), Multilingualism, Second Language Learning, and Gender.
Berlin/New York: Mouton De Gruyter, pp. 17–52.
Pavlenko, A. and Piller, I. (2007), ‘Language education and gender’, in May, S. (ed.)
The Encyclopedia of Language and Education, volume 1. Springer: New York, pp. 57–69.
Pennycook, A. (1994), The Cultural Politics of English as an International Language.
London: Longman.
Pennycook, A. (2007), Global Englishes and transcultural flows. London/New York:
Routledge.
Phillipson, R. (1992), Linguistic Imperialism. Oxford: Oxford University Press.
Piller, I. (2002), Bilingual Couples Talk: The Discursive Construction of Hybridity. Amsterdam:
Benjamins.
Piller, I. (2007), ‘Cross-cultural communication in intimate relationships’, in Kotthoff,
H. and H. Spencer-Oatey (eds), Intercultural Communication. Berlin/New York:
Mouton de Gruyter, pp. 341–359.
Piller, I. and Pavlenko, A. (2004), ‘Bilingualism and gender’, in Bhatia, T. and
W. Ritchie (eds), The Handbook of Bilingualism. Oxford: Blackwell, pp. 489–511.
Piller, I. and Pavlenko, A. (2007), ‘Globalization, Gender, and Multilingualism’, in
Decke-Cornhill, H. and L. Volkmann (eds), Gender Studies and Foreign Language
Teaching. Tűbingen, Germany: Gunter Narr Verlag, pp. 15–30.
Piller, I. and Takahashi, K. (2006), ‘A passion for English: Desire and the language
market’, in Pavlenko, A. (ed.), Bilingual Minds: Emotional Experience, Expression, and
Representation. Clevedon: Multilingual Matters, pp. 59–83.
Polanyi, L. (1995), ‘Language learning and living abroad: Stories from the field’, in
Freed, B. (ed.), Second Language Acquisition in a Study Abroad Context. Amsterdam/
Philadelphia, PA: John Benjamins Publishing Company, pp. 271–291.
Pujolar, J. (2007), ‘Bilingualism and the nation-state in the post-national era’, in Heller,
M. (ed.), Bilingualism: A Social Approach. Hampshire/New York: Palgrave, pp. 71–95.
Raijman, R., Schammah-Gesser, S. and Kemp, A. (2003), ‘International migration,
domestic work, and care work: Undocumented Latina migrants in Israel’, Gender and
Society, 17, (5), 727–749.
Roy, S. (2003), ‘Bilingualism and standardization in a Canadian call center: Challenges
for a linguistic minority community’, in Bayley, R. and S. Schecter (eds), Language
Socialization in Multilingual Societies. Clevedon: Multilingual Matters, pp. 269–287.
Schlieben-Lange, B. (1977), The language situation in Southern France. Linguistics, 19
101–108.
Schüpbach, D. (2005), ‘Shared languages, shared identities, shared stories? A quali-
tative study of life stories by immigrants from German-speaking Switzerland in
Australia’, Unpublished doctoral dissertation, Department of Linguistics and
Applied Linguistics, University of Melbourne.
Seabrook, J. (2007), Cities. London: Pluto Press.
Smagulova, J. (2008), Kazakhstan: Transformation of Language Ideology. Paper presented

at the 17th Sociolinguistics Symposium (SS17), April, Amsterdam, Netherlands.
Sonntag, S. (2003), The Local Politics of Global English: Case Studies in Linguistic Globali-
zation. Lanham, MA: Lexington Books.
Spedding, A. (1994), ‘Open Castilian, closed Aymara? Bilingual women in the Yungas
of La Paz (Bolivia)’, in Burton, P., Dyson, K. and Sh. Ardener (eds), Bilingual Women:
Anthropological Approaches to Second-Language Use. Oxford and Providence, RI: Berg,
pp. 30–43.
Takahashi, K. (2006), Akogare and English Language Learning: Japanese women in Australia.
PhD, University of Sydney.
Tokuhama-Espinosa, T. (ed.) (2003), The Multilingual Mind: Issues Discussed by, for, and
about People Living with Many Languages. Westport, CT and London: Praeger.
Vitanova, G. (2004), ‘Gender enactments in immigrants’ discursive practices: Bringing
Bakhtin to the dialogue’, Journal of Language, Identity, and Education, 3, (4),
261–277.
Watts, J. (2005), ‘Mother tongue’, Mail & Guardian Online, 2005-12-07, http://www.
mg.co.za/article/2005-12-07-mother-tongue.
Williams, E. (2006), Bridges and Barriers: Language in African Education and Development.
Manchester: St. Jerome Publishing.
CHAPTER
2
Language and Economy
Florian Coulmas
2.1 The Economic Turn
One of the effects of the capitalist order that started evolving in Europe some five
hundred years ago and now encompasses the entire globe is that more and more
aspects of social life are being made subject to evaluation and market principles.
The imperatives of profit maximization, competition and capital accumulation
have fanned out from the narrow confines of business activity to many other
areas of human behaviour, as attested, for instance, by the commercialization of
sport and religion. Given the overall importance of language for human behav-
iour, thought and interaction, it is not surprising that language too has become
an object of economic deliberation on the part of both economic actors and econ-
omists. At the same time, linguists have increasingly taken notice of the fact that
languages have certain systematic properties that lend themselves to analysis by
means of economic theory and that language behaviour follows principles, some
of which concur with principles assumed in economic models. Moreover, sociolo-
gists of language acknowledge the importance of economic variables in properly
analyzing the distribution of languages and their varieties in societies across the
globe.
There are hence several points of contact between the economic and language
sciences. Linguists and economists both attempt – for different reasons – to uncover
the underlying regularities of language and language behaviour in order to under-
stand why languages are as they are. Ever since Ferdinand de Saussure, and with
renewed emphasis since Noam Chomsky, theoretical linguistics has tried to stake
out its proper territory, emphasizing its prerogative to determine what language
is and how it should be studied. This however has not prevented the appearance
of several hyphenated sub-disciplines, such as neuro-, psycho-, socio-, ethno- and
geo-linguistics, all with their specific interest in and approach to the study of lan-
guage. This variety reflects the centrality of language to human existence. Unlike
the fields just mentioned, ‘econo-linguistics’ is not a recognized sub-discipline of
the language sciences, nor is there a unified body of knowledge that warrants des-
ignation as ‘economics of language’ with its own journals and scholarly societies
such as, for example, the sociology of language. Yet, thanks to the breadth of inter-
est that economists take in human behaviour and the pervasive nature of language
Language and Economy 29
in all human affairs, the common concerns of economists and linguists are
manifold. The purpose of this article is to review the most important ones.
2.2 Economics and Language
Economists take an interest in language for various reasons. Two broad areas can
be distinguished: language in economics and language in the economy. Economists
produce theories and models that are verbalized in texts and discourses, while eco-
nomic actors use language for a range of purposes, all related in one way or another
to the overarching objective of revenue maximization and cost minimization.
These two fields are reviewed individually below.
2.3 The Language of Economics
2.3.1 The Language of the Profession
Like astronomers, biologists and palaeontologists, economists are concerned with

their professional language as a matter of necessity because claims to knowledge
can be made only by using language (Henderson et al., 1993). Since economists
have been trained to pay heed to efficiency, they perhaps do so more than other
scientists. In economic models, interpretation is an essential ingredient not just
with respect to logical consistency, but also because economic theories are at risk
of being distorted when communicated to the public. Since the 1980s, much dis-
cussion about the methodology of economics has revolved around the rhetoric
and style of economic writing. Behind the ensuing interest by economists in ana-
lytic philosophy, semantics and pragmatics lies a fundamental critique of economic
theory building, which, for the sake of the model, has often treated information as
objective. The economists’ concern with their own professional language can thus
be understood as a response to the emphasis on the social construction of know-
ledge by philosophers such as Derrida (1967) and Rorty (1979). Economics is the
study of choices by actors who are credited by a large faction of economists with
‘rationality’ (Levy, 1997). The outcome of any choice is dependent on the infor-
mation available to the agent. Treating information as objective, rather than being
an excusable simplification, is a serious flaw calling for correction. During the past
two decades, a great deal of discussion about economic theory and methodology
has dealt with bounded rationality and imperfect information. Rubinstein (2000),
among others, suggests that discourse analysis, pragmatics and the philosophy of
language provide useful tools for a meta-critique of economics in this spirit. He
argues, for instance, that the rhetoric of game theory, a major tool of economic
modelling, is misleading because it not only creates an illusion of preciseness but
because it also suggests that the notion of ‘strategy’ in game theory can be inter-
preted as a course of action in the real world, which is not the case. For Rubinstein
this is a major reason why economists should study the language of economics.
30 Florian Coulmas
Taking this line of thought further, some economists have identified the power
of language as a major force determining the success or failure of economic theo-
ries. To these economists, linguistic determinism, especially the ‘metaphors-we-
live-by’ strand, has a strong appeal (Gramm, 1996; Levy, 1997). The metaphors
of economics, they hold, can influence behaviour and attitudes, including those
of the scientific community. As Ferraro et al. (2005: 21) put it in a discussion of the
effects of language on the social categories that people use to interpret their real-
ity, ‘theories become dominant when their language is widely and mindlessly used
and their assumptions become accepted and normatively valued, regardless of
their empirical validity.’ The word ‘mindlessly’ is critical for the argument, because
language is also the instrument used to expose the lack of a theory’s empirical
validity, but the authors clearly have good reasons for attacking both the uncritical
acceptance of terms and the conceptual frameworks that come with them.
2.3.2 Optimizing Language
Interest in the professional language of economics is coupled with a more general

interest in how natural language works and how it has come to work as it does.
Evolutionary economics looks at language as the result of a long-term process
in which an ‘invisible hand’ seeking optimality is at work. Marschak (1965) first
developed the notion that the minimization of the ‘cost’ of communication is a
force in the evolution of language. This idea, which emphasizes the functionality
of language, has been carried further in evolutionary game theory (Maynard
Smith, 1982). The basic assumption is that language behaviour consists of a set of
conventions shared by all members of the community and that if a language does
not fulfil its community’s needs, some ‘mutant’ members will expand or otherwise
deviate from existing conventions to improve its functioning (Rubinstein, 2000: 27).
However, with a system as complex as natural language, it is not easy to determine
what is and what is not functional. Models of evolutionary game theory are stripped
of most complexities; all terms must be unequivocally defined. Meanwhile, econo-
mists in an attempt to put their professional language in order are forced to deal
with the relationship between concepts and words in the real world. Vagueness is
one of the most striking features of this relationship in everyday speech. It is also
inherent in many economic variables (such as poverty, satisfaction and prefer-
ence), as well as in the terms designating these variables. Both the terms and the
concepts are vague and contextual. Not surprisingly, therefore, economists have
paid attention to what linguists and philosophers of language have had to say about
the prevalence of semantic underspecification in natural languages (e.g. Qizilbash,
2003).
2.3.3 Interpretation and Rhetoric
A related aspect of language that economists have tackled is how language behav-
iour affects action (Samuels, 1990). On this issue too, there is the economist’s
perspective and the economic actor’s perspective. Both the former and the latter
habitually spend time and effort to communicate their ideas with the purpose of
influencing the beliefs and behaviours of their audience. In this endeavour, the
truth of a statement and its plausibility play a critical role. The importance of rhet-
oric in this context is the subject of much of McCloskey’s (1983) work, upon which
others have built (e.g. Klamer, 2001, Mäki, 2000). This is important because much
of economic modelling is about decision making based on beliefs about the world.
How non-arbitrary decisions are possible under conditions of vague terms and
incomplete information is a major challenge not just for rational choice theory,
but also for the economic activities it is intended to explain, as Hayek (1945) had
already pointed out ahead of his time. The relationship between the state of the
world and its linguistic representation, i.e. how people can make sense of it and
how they use words to convey their meaning, are, therefore, questions of major
concern to economists. Economic models strive for mathematical exactitude, but
unlike mathematical models they are tied to the real world – no matter how
flimsily – and thus require linguistic interpretation. Economics cannot rid itself
of this soft spot, which accounts for its vacillating nature between the humanities
and the sciences.
No one has pushed economics more forcefully in the direction of the humani-
ties than McCloskey, who has said of economists that they are poets without know-
ing it (McCloskey, 1990: 12), and has constructed a theory of economic activity as
conversation. Language is of such sweeping importance in the economic process
that the entire economy is best characterized as conversation. Accordingly, eco-
nomics must be reconstructed as rhetoric – the analysis of persuasive talk which
enables inventors to raise capital, entrepreneurs and labour to negotiate and coop-
erate, advertisers to appeal to customers, stockbrokers to talk to clients and to each
other and so on (McCloskey, 1994: 373–374). In short, economic actors talk with
each other, a fact that has not been lost on economists. As Alvesson and Kärreman
(2000: 147) put it, ‘most qualitative studies of organizations focus on talk’. Apart
from pre-capitalist forms of silent barter (Calvet, 1984), the market is a place of
exchange: the exchange of goods, services and, increasingly, knowledge, mediated
by linguistic exchange. The reliability of the medium of exchange and a proper
understanding of its logic, the logic of conversation, are crucial for the smooth
working of economic transactions. The ‘cooperative principle’, proposed by philos-
opher of language Paul Grice (1975) as the basis of verbal interaction, has there-
fore been accepted by economists (e.g. Sally, 2004). Language is built on trust; i.e,
on the premise that people mean what they say and that they speak the truth – or
at least what they believe to be true. The same holds for money as the means of
economic exchange, in that money has no intrinsic value and presupposes mutual
trust (Ingham, 2006).
Other theories McCloskey and others have drawn on in their analysis of eco-
nomic texts and in their attempt to assess the role of language in the economy
include literary theory, rhetoric and speech act theory. As mentioned, McCloskey’s
concern is not limited to a critical appraisal of the language of economists, for
language plays a vital role in the economy too: simultaneously in planning and
32 Florian Coulmas
executing transactions, as a means of production, and as a product. In short, lan-

guage is a driving force of the economy to be reckoned with. McCloskey and
Klamer (1995) have gone so far as to put a cipher on the importance of language
in the economy. According to their estimate, persuasive speech acts account for
some 25 per cent of US GNP. Although this would seem rather hard to verify, the
merit of McCloskey’s approach is to focus attention on language not just as a tool
that economists use, but as a phenomenon that affects economic processes in mul-
tiple ways that economists cannot afford to ignore.
2.4 Economics of Language
When analysing the distribution of wealth in modern societies, economists have

traditionally divided economic actors into producers and consumers, as well as
into social classes, where class is defined either in Marxian terms of control over
the means of production, or along Weberian lines relating to the assets which indi-
viduals bring to the market and which determine their life chances. Assets consid-
ered in this context include property, education, parents’ occupation and income,
among others. Language has not usually entered the picture as a determinant of
class or a variable in resource allocation. It was only when in advanced capitalist
societies language became an issue of community relations that its potential impact
on economic processes came into view. Canada, a multicultural country with two
official languages, played a pioneering role in this realization.
2.4.1 Language Skills as Human Capital
A growing separatist movement in Quebec, where French is the majority language,

and the discovery that competence in only the French language was associated
with lower income (Migué, 1970; Durocher and Linteau, 1971; Breton and
Mieszkowski, 1977) triggered a range of studies about the economic utility of
languages and their effects on earnings (reviewed in Christofides and Swidinsky,
1998). Research about the relationship of language and income distribution was
also carried out in other multilingual countries such as the United States (Bloom
and Grenier, 1996; Chiswick, 1978) and Switzerland (Grin, 1999), among others.
These studies incorporated language in human capital theory.
Human capital theory is concerned with the productive power of labour, as
distinct from other forms of capital, and serves as a major instrument to explain
income differences. In abstract terms, learning a language can be considered an
investment of resources – tuition fees, time, learning materials – in human capital
(Becker, 1975; Riedler and Pons-Riedler, 1986; Robinson, 1988). In this regard,
language is similar to other skills that can be acquired through learning and training.
What makes language as a component of human capital interesting is that both
the ‘inherited capital’ of actors’ first languages and that of subsequently acquired
languages does not yield the same rate of return on investment. The same is true
of resources allocated to improving language skills on both an individual and

collective level (Breton, 1978, 1998a, 1998b). This means that, for instance, an
investment in learning Choni, a Sino-Tibetan language of some 24,000 speakers, is
likely to yield a lesser return – if any – than learning Chinese. As a component of
human capital, competence in Choni and Chinese would accordingly be evaluated
differently, to cite a rather extreme example. However, rates of return are likely to
differ at any given place and time even for more balanced language pairs. This has
complex implications for calculating the contributions of specific language skills
to human capital. Languages are local. Competence in Dutch, for instance, is less
valuable for a Moroccan living in the Atlas Mountains than for one who lives in
Amsterdam or for one who lives in Casablanca working for a trading company that
does business with the Netherlands. The evaluation of languages as components of
human capital more generally is even more complicated.
Most economists subscribe to the notion that cost efficiency is a guiding princi-
ple of human action. Regarding language as human capital, this tenet leads to
the hypothesis that actors will prefer to invest time and money to acquiring or
improving languages that, in the context in question, yield the highest return.
Though intuitive and to some extent borne out by observation, this hypothesis is
hard to substantiate because of the semantic under-determinacy of ‘return’. The
double nature of language as tool and symbol accounts for the difficulty. In a
study on the costs and benefits of language policies in Quebec, Vaillancourt (1985,
1987) therefore proposed a methodology that makes a distinction between quan-
tifiable costs and non-quantifiable ‘psychic’ costs. Analogously quantifiable and
non-quantifiable returns on investment must be reckoned with when drawing up
models of what motivates individuals to make language investments.
If human capital theory is to deal with language, psychic or symbolic expendi-
tures and returns must be reckoned with. While little is known about the empirical
weight of this type of expenditure and return, as compared to quantifiable costs
and benefits, they clearly add a level of complexity to any model designed to map
decisions about language investment. Robinson (1988) presented such a model
for a simple economy in which individuals are endowed with one of two possible
languages, assuming that ‘households choose the language that maximizes the
joint utility of the members’ (Robinson, 1988: 56). Because of many other assump-
tions, which Robinson feels forced to make in order not to let the complexity of
the model get out of hand, his model is as far removed from empirical observation
as many other economic models, removed because (1) speakers of the two lan-
guages are uniformly distributed over the area of the economy, and (2) individuals
are not allowed to be immigrants. If language change occurs, it does so at the
instant of household formation. Given that immigration is one of the most con-
spicuous situations where language choice and economic activity interact, these
and some additional assumptions make it highly unlikely that the model is suitable
‘to test hypotheses of language acquisition as part of an optimal human capital
investment’ (Robinson, 1988: 54), as it is intended to do. To Robinson’s credit it
must be said that together with other Canadian scholars he has helped to advance
the notion that decision theory is applicable to language choice.
34 Florian Coulmas
2.4.2 Differential Value of Languages
Because of the symbolic function of language as a marker of identity, the notion of

language evaluation against a given standard has often been rejected on moral
and/or political grounds. Yet, the instrumental side of language lends itself to eco-
nomic valuation (Vaillancourt, 1991; Coulmas, 1992). Various scales of measure-
ment have been proposed, notably the numerical strength of a speech community,
its GDP and the number of speakers of other languages that use or study the lan-
guage as a foreign language (Ammon, 2003). Relating GDP and language illumi-
nates one aspect of the economic significance of different languages. Tabulating
data of the percentages of world GDP by languages for the years 1975 to 2002,
Davis (2003) ranks the 18 most widely spoken languages of the world. Not surpris-
ingly, English accounted for the largest share of world GDP in 2002 with 29.3 per
cent, followed by Chinese with 12.5 per cent and Japanese with 7 per cent. Multi-
lingualism in many parts of the world obscures the picture and in some cases –
Hong Kong for example – makes it difficult unequivocally to assign a language to
an economy. Yet, relating languages and GDP should not be rejected prematurely.
Since Davis’ analysis includes a timeline, he is able to point out a number of trends,
such as the steady rise of Chinese and the relative decline of Japanese and virtually
all European languages. It is also worth noting that the rest of the languages of
the world taken together account for just 12.5 per cent of world GDP. However
accurate the information used in these calculations, the resulting picture is of
interest both in terms of the levels of economic activity in different languages
and the survival chances for economically insignificant languages in a world
increasingly subservient to the imperatives of utility, profit maximization and capi-
tal accumulation.
It must of course be kept in mind that no value scale allows for an absolute
rating of the languages of the world. Every evaluation must be indexed for time,
place and purpose. Yet, assuming that a uniform and well-defined standard of util-
ity of a language becomes available, it is conceivable that considerations along
these lines can be fruitfully combined with evolutionary economics to study the
world system of languages.
2.4.3 Language Choices
Languages are both inherited and chosen by individuals and governments

(Coulmas, 2005). Like the former, the latter are influenced in their choices by
instrumental as well as symbolic concerns. Nationalism provides the motivation to
choose – i.e., legally protect, institutionally support and financially promote – a
certain language for national and/or official functions (often the default choice).
By contrast, the choice of languages for foreign language curricula is more likely
to be influenced by instrumental considerations of utility. These choices are typically
made with a view on what policy makers think is in the best interest of the nation,
while the effects of their choices for the languages concerned are not considered.
However, thanks to the network character of language there are such effects that
lend themselves to yet another economic theory: the network externalities theory
(Katz and Shapiro, 1985).
Joining a network is beneficial for every new entrant, for example to gain access
to services or information. At the same time, a network becomes more useful and
hence valuable as it gains new participants. This phenomenon is called network
externalities. Examples of this effect include telephony, transport systems and the
internet. What is the use of internet access if few other people have it and little
information can be acquired through it? The network must have a certain size to
be useful, and every new entrant benefits from joining, but also makes it a little
bit more useful for everyone else. The law of diminishing returns applies here.
One-hundred-thousand extra or fewer speakers make little difference to Chinese
but a big difference to Choni. Network technologies are furthermore associated
with indirect externalities since the necessary equipment becomes cheaper as the
number of users grows. Languages can be considered networks, and learning a
foreign language means gaining access to a network, i.e access to all individuals
who speak the language and the information stored in it. Again, indirect externali-
ties can be identified. Dictionaries and other reference works for Choni will always
be more expensive than for Chinese because the network of potential users from
whom the cost of producing these materials can be recovered is so much smaller.
While language learners do not usually take external effects of their joining the
network into account, such effects are incontrovertibly there. Every new member
increases the human capital value afforded by competence in the language in
question. This dynamic has a potentially positive feedback loop: The bigger the
network of speakers, the greater the utility of the language; and the greater a
language’s utility, the more likely it is to attract new speakers. Whether there is
a threshold and, if there is, whether it is quantifiable, are interesting questions
waiting to be studied (Grin, 1993). From the point of view of the language and the
learners concerned, positive feedback loops are generally regarded as positive
externalities. Negative externalities are however also conceivable if we change
our point of view. Again, the tension between the practical-instrumental and the
cultural-symbolic side of language comes to bear here. As Dalmazzone (1999) has
pointed out, joining a speech community implies exposure to that community’s
culture and ideologies which may have corrosive and even destructive effects on
other communities. This perspective is of particular relevance in minority situa-
tions, for while learning a language does not imply unlearning another, the lan-
guage of greater utility may well drive that of lesser utility out of the market.
Foreign language curricula are not generally designed with a view on network
externalities. Rather, much like individuals, education ministries make language
choices based on perceived needs and expected utility. Providing a minority access
to a big language is seen as beneficial to that community, while the benefits gener-
ated for the big language community are ignored. This is because in this context
language is taken into account from an instrumental point of view only while its
cultural and symbolic functions for community integration and identity are ignored.
The distribution of European languages as official languages in the post-colonial
36 Florian Coulmas
world is a testimony to this dynamic. The bigger language is the more useful one.
The size of the speech community is not the sole determinant of a language’s util-
ity, but it is an important one, if only because it is considered coterminous with the
size of the market (Hocevar, 1983).
The idea that languages delimit markets was already discussed at the beginning
of the twentieth century when the ‘Powers’ were busy dividing the world among
themselves and before the language of the British Empire was universally recog-
nized as the uncontested front-runner. Quoting Vladimir Lenin is a bit out of
fashion these days, but most economists will concur with him that ‘the uniformity
of language and its unimpeded development is one of the most important presup-
positions of a truly free and all-encompassing trade commensurate with modern
capitalism’ (Lenin, 1980: 398). A big market promises higher profits; the question,
however, is to what extent uniformity of language is conducive to, or engendered
by trade.
2.4.4 The World Language Problem
Lenin thought that a common language was a prerequisite for modern capitalism.
By contrast, the notion that the level of economic development determines the
level of linguistic diversity in a country has also been advanced (e.g. Greenberg,
1956; Gellner, 1983). How economic development and linguistic homogeneity
are related is a question of interest to economists and language planners alike
(Grin, 1996). Two influential studies on this correlation by Fishman (1968) and
Pool (1972) found that linguistic fragmentation is detrimental to economic devel-
opment. Fishman (1968: 60) compared a group of linguistically heterogeneous
countries with a linguistically homogeneous group and concluded that the latter is
related to many more of the positive indices of development – high life expect-
ancy, low infant mortality, high GDP and GDP per capita – than the former. And
Pool (1972: 225), on the basis of a similar comparison, found that while linguisti-
cally relatively homogeneous countries could be poor, heterogeneous ones could
never be rich. Nettle (2000) has re-examined these findings using data from the
World Bank World Development Report 1993. He found ‘some evidence of an inverse
relationship between linguistic heterogeneity and the level of economic develop-
ment’ (Nettle, 2000: 344). However, he did not judge the relationship strong
enough to recommend enforcing linguistic homogeneity as a remedy to poverty,
quite apart from the ethical questions such a recommendation might raise. In the
long run, the economic situation is likely to determine the linguistic one, for as
globalization continues, so does the loss of human languages (Nettle and Romaine,
2001). This view is tilted to the instrumental function of language and appeals
to free-market ideologues who are happy to assume that the market will find the
optimal answer to the question of how people should communicate with each
other. By contrast, defenders of minority rights – who, together with the instru-
mental function of language, emphasize its symbolic function for identity and rec-
ognize diversity as inherently valuable – conceptualize multilingualism as an asset
deserving of state protection (Walsh, 2006).
Isolating the factor of language remains difficult, and an unequivocal correla-

tion between multilingualism and economic development has proven hard to
establish, because the wealth of nations is dependent on so many variables. Yet, in
recent decades, the capitalist market economy has reached the remotest parts
of the world. And as outlying areas have been more tightly integrated into the capi-
talist system, the more readily small speech communities have moved away from
their languages because parents see better life chances for their children in lan-
guages associated with commerce and business opportunity. Maximizing one’s
communication opportunities to improve employment chances is a strong force
behind language shift, a process that seems to be accelerating in the wake of glo-
balization and which works to the detriment of small languages. Clearly, European
colonialism and then American military and economic hegemony have been the
main historical developments that have lead to a situation in which the only place
left for many small languages is in the archives of linguists determined to preserve
a part of human heritage that has lost its immediate utility.
2.4.5 The Commodification of Language
Even as small languages disappear, there is a growing demand for language skills,
driven by a number of factors: the need and desire to participate in the ever more
integrated world economy; cross-border traffic of goods, services and people; an
increasingly mobile reserve army of migrant labour; mass tourism; and the com-
munications revolution. Language itself has become the mainstay of a multibillion
dollar industry. Selling language is big business. The economic inequality of
languages finds vivid expression in the language market. The vast majority of the
world’s languages don’t sell at all, while a handful of European languages aug-
mented by a few Asian ones generate enormous revenues, English being ahead of
the pack by a large margin. A 1989 Economist Intelligence Unit report entitled
English: A World Commodity assessed the world market for English programmes,
textbooks and other materials at between £6.0 billion and £12.0 billion in 1988
(McCallen, 1991).
Almost all English-speaking countries compete for the profits this huge market
generates through publishers, teacher training programmes, language schools,
holiday language courses, etc. Because of the multifaceted nature of national EFL
markets and the many indirect profits they yield, reliable statistics are scarce and
estimates on the size of the world market accordingly difficult. One thing is
certain: since McCallen’s study was published this market has expanded further,
especially in Eastern Europe and China (Hildebrandt and Liu, 1991) after Soviet
political influence waned. It is obvious then that an economic valuation of lan-
guages not only takes place but also has a transforming effect on the linguistic
situation of the world, and that English – driven by the demand from hundreds
of millions of people around the globe – is assuming the function of a universal
second language.
It should be noted, however, that during the past quarter century English has
not been the only market to expand. Rather, as world trade increased, demand
38 Florian Coulmas
rose also for other languages associated with strong economies, such as German,
French, Japanese and, of late, Chinese. The accelerating spread of English not-
withstanding, knowledge of the local language is widely considered a business
advantage (e.g. Fixman, 1990; Metcalf, 1992; Gauthieret al., 2007). Web-based
dictionaries, translation assistance software and other advanced language learning
tools have become the vanguard of a knowledge-intensive industry.
2.4.6 Terms of Trade
During the past quarter century, the volume of world trade has increased signifi-
cantly relative to world output (Legrain, 2006). This means that trade has become
more important in generating wealth. A common language facilitates trade
between individuals and companies (Lazear, 1999). Language boundaries by con-
trast function as trade barriers, a view that is commensurate with the notion that
language diversity means increased transaction costs. In this regard, languages
exhibit functions that are reminiscent of currencies. Both are media of exchange,
without which complex transactions cannot take place. As early as the eighteenth
century, the Scottish philosopher and economist David Hume (1964: 263) identi-
fied the heart of this analogy as ‘a human convention without any promise’, dwell-
ing on the same crucial property of language mentioned above in connection
with the ‘cooperative principle’ of conversation and its significance for economic
theory, namely trust. The functional similarity of language and money is not just
a metaphorical one. Its many intriguing aspects have been explored at length by
one of the founding fathers of sociology, Georg Simmel (1900). A historical over-
view of this analogy, which has been remarked upon many times, can be found in
Coulmas (1992, chapter 2). In trade, currencies and languages function similarly.
The effect of replacing several national currencies by a single currency – as the
Euro in the European Economic and Monetary Union – is to increase trade and
to reduce transaction costs. Having a common language likewise enhances trade,
whereas the absence of one has often led to the creation of a trade language. The
role of language in trade has been documented in several detailed studies, for
example: Cremer and Willes (1991) on Macau; Holden (1987) on British firms’ activ-
ities in France, the USSR and Japan; Guy (1992) on foreign language competence of
US companies; and Marschan-Piekkari et al. (1999) on a Finnish multinational. The
gist of these studies is that insufficient language competence is detrimental to
trade and costs companies dearly.
2.4.7 Reducing Transaction Costs
With increased globalization of business and the economy, organizations have

become more aware of the importance of communication competence, including
language. How to reduce the transaction costs associated with linguistic diversity
is a genuinely economic question that applies to cross-linguistic communication
both within companies and with their external environment. Companies need to
be able to rapidly and correctly interpret information potentially relevant to their

strategies and must secure smooth communication between management and
personnel, between headquarters and subsidiaries, as well as with their customers,
wherever they are. Should a company decide on a single ‘company language’?
Should it provide language training for its personnel, employ translators and inter-
preters, buy automatic translation software, or outsource language work? More
and more internationally active businesses have to find answers to these questions,
answers that will meet their specific needs. Failure to do so can be costly. Econo-
mists have therefore also begun to take notice of the importance of language
choices for business organizations. Dhir and Savage (2002) reviewed a number of
economic theories and frameworks that may be applied to solving the language
problems of organizations and proposed a model based on social judgment theory
for assessing the value of a working language. Their model is rather complex,
involving five evaluation parameters as well as a management committee assigned
the task of developing a policy to select a working language most suitable for the
company. This complexity reflects some of the cultural, linguistic, demographic
and social factors that have a bearing on language choice, although the actual
decision processes of business organizations are likely to be even more involved.
The model shows that reducing language-induced transaction costs is a compli-
cated task, but one that organizations can ignore only at a price. The model offers
one interpretation of the notion of the economic value of a language relative to
specific purposes and thus contributes to a better understanding of the role lan-
guage plays in the creation and management of organizational capital.
2.4.8 Language as a Public Good
For trading partners it is beneficial to share a common language, as indeed it is

beneficial to anyone who decides to use that common language for the purposes
of trade. There is no secret access code; no one can be excluded from obtaining
benefits from its use. And there is no good reason to want to exclude anyone from
using it, for one actor’s use does not reduce the amount that is left for others.
These are the defining features of public goods: non-excludability and non-rivalry.
As the network-externalities approach to languages discussed above makes clear,
languages, though valuable, are not used up by being employed by additional users
(speakers). On the contrary, additional users increase a language’s utility. Language
use is not consumption of exhaustible supplies, and languages can, therefore, be
conceptualized as public goods.
Economists have not usually looked at language in this way, because it is consid-
ered a natural endowment that defines humanity itself. At the same time, every
language is an artefact, albeit one that is neither well planned nor consciously
designed as a whole. Just as cities grow as people add one house after another with-
out following an overarching blueprint, languages are artefacts that have come
into existence without a general design. New words are coined every day, most
of them because the expressive resources of the language in question are found
wanting. Every word of every language was used at some point for the first time.
40 Florian Coulmas
When a new word proved useful it was adopted and became part of the public
good that is the community language.
The existence of dozens of language academies around the world lends
additional support to the notion that languages are public goods that serve their
communities. These academies are institutions charged with the preservation,
maintenance, development and distribution of the public good of a national lan-
guage or languages. An example is the Pan South African Language Board, which
is responsible for the country’s eleven official languages. The art of writing has
made language an even more valuable tool, bringing its nature as a public good
into sharper focus. Deliberate cultivation and promotion of this tool by academies,
schools and other institutions testify to the value attributed to it. Its value is depend-
ent on its systematic properties, but also – , and critically – on the fact that it is a
public good. This is evidenced by languages such as Akkadian and Egyptian, lan-
guages that are as rich and elaborate as any, yet have lost their speech communities
and with them their value as a public good.
Positing language as a public good opens up several interesting avenues of
research yet to be explored. Some of these are listed below.
1. Public goods do not necessarily come free of charge. While everyone benefits
from public goods – and in some cases cannot live without them – not everyone
is willing to pay their share to maintain them. Can the free-rider problem be
meaningfully interpreted with respect to language?
2. The public good nature of languages rests in their being artificial tools. Are
there superior and inferior tools? Changes on the linguistic map of the world
are commonly explained in terms of social, political and economic variables
acting on populations. Could the intrinsic properties of the linguistic tools that
communities have at their disposal be a part of the equation?
3. It is well known that both incipient languages (emerging pidgins) and mori-
bund languages spoken by shrinking communities are characterized by reduced
expressive power. Both kinds of languages are found in language contact situa-
tions only where competing languages assume some of the functions of the
public good. Is it desirable and feasible to influence this division of labour?
4. Because of the free-rider problem, public goods call for public policies. In view
of the expected gains of enhanced trade and access to a wider knowledge base,
the diffusion of a common language has often been supported by public policy,
for instance in countries that promote English as a second language through
their school system. Under what circumstances can minority languages obtain
recognition as a public good deserving of protection? The case of Ireland comes
to mind, where Irish is treated as a public good although it is not considered to
have much economic utility for anyone.
5. How do instrumental and symbolic values of language interact in determining
policies designed to intervene in the market of competing languages?
6. Can the case be made that linguistic diversity, like the diversity of species, should
be considered a global public good, deserving of policies that protect that
diversity?
7. Is English becoming a global public good and, if so, should a global policy for
its relation to other languages be developed?
8. What does a public goods approach to language imply for the allocation of
public funds to subsidize language export?
These are some of the questions that a public goods approach to languages
might be employed to address, taking into account the economic characteristics of
language discussed in this chapter, i.e. languages as human capital, as networks, as
transaction costs, as trade barriers, as systems built on trust and as a means of
exchange, as commodities, as tools and objects of public choice and as potential
impediments to economic development. Properly understood, these are economic
aspects pertaining to language use patterns. Other questions also arise regarding
the internal organization of language as a highly complex symbolic system – ‘lin-
guistic capital’ – resulting from many generations’ collective and cooperative
‘linguistic work’ (Rossi-Landi, 1975). As tools are put to use following the ‘principle
of least effort’, they evolve, become simpler and improve (Zipf, 1949). Discovering
those structural features of language that testify to the working of the optimality-
seeking invisible hand is a worthy project in itself, though one that must be the
subject of another chapter.
References
Alvesson, M. and Kärreman, D. (2000), ‘Taking the linguistic turn in organizational

research: Challenges, responses, consequences’, The Journal of Applied Behavioral
Science, 36, (2), 135–158.
Ammon, U. (2003), ‘The international standing of the German language’, in Jacques
Maurais and Michael A. Morris (eds), Languages in a Globalising World. Cambridge:
Becker, G. S. (1975), Human capital. A Theoretical and Empirical Analysis with Special
Reference to Education. New York: Columbia University Press. (Précis: http://www.
econlib.org/library/Enc/HumanCapital.html), accessed 11.09.2009.
Bloom, D. E. and Grenier, G. (1992), ‘Economic perspectives on language: The rela-
tive value of bilingualism in Canada and the United States’, in James Crawford
(ed.), A Source Book on the Official Language Controversy. Chicago: University of
Chicago Press, pp. 445–452.
Bloom, D. E. and Grenier, G. (1996), ‘Language, employment and earnings in the
United States: Spanish-English differentials from 1970 to 1990’, International Journal
of the Sociology of Language, 121, 45–68.
Breton, A. (1978), ‘Nationalism and language policies’, Canadian Journal of Economics,
11, 656–668.
Breton, Albert (1998a), ‘An economic analysis of language’, in Albert Breton (ed.),
Economic Approaches to Language and Bilingualism. Ottawa: Department of Public
Works and Government Services Canada, pp. 1–33.
Breton, A. (ed.) (1998b), Economic Approaches to Language and Bilingualism. Ottawa:
Department of Public Works and Government Services Canada.
42 Florian Coulmas
Breton, A. and Mieszkowski, P. (1977), ‘The economics of bilingualism’, in W. E. Oates

(ed.), The Political Economy of Federalism. Lexington, MA: Lexington Books,
pp. 261–273.
Calvet, L.-J. (1984), ‘Troc, marché et échange linguistique’, Langage et Société, 27, 55–81.
Chiswick, B. (1978), ‘The effect of Americanization on the earnings of foreign-born
men’, Journal of Political Economy, 85, 897–921.
Christofides, L. N. and Swidinsky, R. (1998), ‘Bilingualism and Earnings: A Study Based
on 1971, 1981 and 1991 Census Data’, in Albert Breton (ed.), Economic Approaches
to Language and Bilingualism. Ottawa: Department of Public Works and Services
Canada, pp. 121–185.
Coulmas, F. (1992), Language and Economy. Oxford: Blackwell.
Coulmas, Florian (2005), Sociolinguistics. The Study of Speakers’ Choices. Cambridge,
New York: Cambridge University Press.
Cremer, R. D. and Willes, M. J. (1991), ‘Stock in trade: The role of English in interna-
tional trade proceedings’, Journal of Asian Pacific Communication, 2, (1), 103–116.
Dalmazzone, S. (1999), ‘Economics of language: A network externalities approach’, in
Albert Breton (ed.), Exploring the Economics of Language. Ottawa: Department of
Public Works and Government Services Canada, pp. 63–87.
Davis, M. (2003), ‘GDP by language’. Unicode Technical Notes #3: http://macchiato.
com/economy/GDP_PPP_by_language.html (accessed 18 March 2008).
Derrida, J. (1967), L’écriture et la difference. Paris: Seuil.
Dhir, K. S. and Savage, T. (2002), ‘The value of a working language’. International
Journal of the Sociology of Language, 158, 1–35.
Durocher, R. and Linteau, P.-A. (1971), Le retard de Québec et l’inferiorité économique des
Canadiens français. Montreal: Éditions du Boréal.
Ferraro, F., Pfeffer, J. and Sutton, R. I. (2005), ‘Economics language and assumptions:
How theories can become self-fulfilling’, Academy of Management Review, 30, (1), 8–24.
Fishman, J. A. (1968), ‘Some contrasts between linguistically homogeneous and lin-
guistically heterogeneous polities’, in J. A. Fishman, C. A. Ferguson and J. Das
Gupta (eds), Language Problems of Developing Nations. New York: Wiley, pp. 53–68.
Fixman, C. S. (1990), ‘The foreign language needs of U.K.-based corporations’, Annals
of the American Academy of Political and Social Science, 511, 25–46.
Gauthier, V., Didiot-Cook, H. and Scheirlinckx, K. (2007), ‘Language needs in busi-
ness, a survey of European multinational companies’. University of Connecticut,
Department of Economics, Ideas: http://ideas.repec.org/p/ebg/heccah/0725.
html (accessed 18 March 2008).
Gellner, E. (1983), Nations and Nationalism. Oxford: Blackwell.
Gramm, W. S. (1996), ‘Economic metaphors: Ideology, rhetoric and theory’, in J. S. Mio
and A. N. Katz (eds), Metaphors: Implications and Applications. Mahwah, NJ: Lawrence
Erlbaum, pp. 147–170.
Greenberg, J. H. (1956), ‘The measurement of linguistic diversity’. Language, 32,
109–115.
Grice, H. P. (1975), ‘Logic and conversation’, in P. Cole and J. L. Morgan (eds), Syntax
and Semantics, vol. III, Speech Acts. New York: Academic Press, pp. 41–58.
Grin, F. (1993), ‘The relevance of thresholds in language maintenance and shift:

A theoretical examination’, Journal of Multilingual and Multicultural Development, 14,
375–392.
Grin, F. (1996), ‘Economic approaches to language and language planning: An
introduction’, International Journal of the Sociology of Language, 121, 1–16.
Grin, F. (1999), Compétences et récompensenses. La valeur des langues en Suisse. Fribourg:
Presses Universitaires de Fribourg.
Guy, B. S. (1992), ‘Dismanteling self-imposed trade barriers revisited: The role of
language development ten years later’ (ERIC # ED 353793). Education Resources
Information Center: http://www.eric.ed.gov/ERICWebPortal/custom/portlets/
recordDetails/detailmini.jsp?_nfpb=true&_&ERICExtSearch_SearchValue_
0=ED353793&ERICExtSearch_SearchType_0=eric_accno&accno=ED353793
(accessed 18 March 2008).
Hayek, F. A. (1945), ‘The use of knowledge in society’, The American Economic Review,
35, (4), 519–530.
Henderson, W., Dudley-Evans, T. and Backhouse, R. (eds) (1993), Economics and
Language. London and New York: Routledge.
Hildebrandt, H. W. and Liu, J. (1991), ‘Communication through foreign languages:
An economic force in Chinese enterprises’, Journal of Asia Pacific Communication,
2, (1), 45–67.
Hocevar, T. (1983), ‘Les aspects économiques de la dynamique fonctionelle des
langues’. Language Problems and Language Planning, 7, 135–147.
Holden, N. (1987), ‘Language barriers as differential constraints on the international
behavior of firms’, in Humphrey Tonkin and Karen M. Johnson-Weiner (eds), The
Economics of Language Use (Report of the Fifth Annual Conference of the Center for
Research and Documentation on World Language Problems). New York: Center
for Research and Documentation on World Language Problems, pp. 119–137.
Hume, D. (1964), ‘Of the origin of justice and property’, in A Treatise of Human Nature.
Aalen: Scientia, Book 3, 258–273 (reprint of second edition, London, 1886).
Ingham, G. (2006), ‘The nature of money’. American Journal of Sociology, 111, (4),
1227–1229.
Katz, M. L. and Shapiro, C. (1985), ‘Network externalities, competition, and compati-
bility’, American Economic Review, 75, 424–440.
Klamer, A. (2001), ‘Making sense of economists: From falsification to rhetoric and
beyond’, Journal of Economic Methodology, 8, 69–75.
Lazear, E. P. (1999), ‘Culture and Language’, The Journal of Political Economy, 107, (6),
Part 2 (=Symposium on the Economic Analysis of Social Behavior in Honor of Gary
S. Becker), 95–126.
Legrain, P. (2006), ‘Why NAMA liberalisation is good for developing countries’, The
World Economy, 29, (10), 1349–1362.
Lenin, W. I. (1980), ‘Über das Selbstbestimmungsrecht der Nationen’. In: Ausgewählte
Werke. Berlin: Dietz Verlag, 2, 393–463.
Levy, D. M. (1997), ‘Adam Smith’s rational choice linguistics’. Economic Inquiry, 35,
672–678.
44 Florian Coulmas
Mäki, U. (2000), ‘Performance against dialogue, or answering and really answering:

A participant observer’s reflections on the McCloskey conversation’, Journal of
Economic Issues, 34, 43–59.
Marschak, J. (1965), ‘Economics of language’, Behavioral Science, 10, 135–140.
Marschan-Piekkari, R., Welch, D. and Welch, L. (1999), ‘In the shadow: The impact of
language on structure, power and communication in the multinational’, Interna-
tional Business Review, 8, 421–440.
Maynard Smith, J. (1982), Evolution and the Theory of Games. Cambridge: Cambridge
University Press.
McCloskey, D. N. (1983), ‘The rhetoric of economics’, Journal of Economic Literature, 21,
481–517.
McCloskey, D. N. (1990), If You’re So Smart. The Narrative of Economic Expertise. Chicago:
University of Chicago Press.
McCallen, B. (1991), ‘The size of the world market for EFL programmes and materi-
als’, Journal of Asian Pacific Communication, 2, (1), 185–189.
McCloskey, D. N. (1994), Knowledge and Persuasion in Economics. Cambridge: Cambridge
University Press.
McCloskey, D. N. and Klamer, A. (1995), ‘One quarter of GNP is persuasion’, American
Economic Review, Papers and Proceedings, 85, 191–195.
Metcalf, H. (1992), Foreign Language Needs of Business. Falmer: University of Sussex,
Institute of Manpower Studies (Report No. 215).
Migué, J.-L. (1970), ‘Le nationalism, l’unité nationale et la théorie économique de
l’information’, Revue Canadienne d’économique, 3, 183–198.
Nettle, D. (2000), ‘Linguistic fragmentation and the wealth of nations: The Fishman-
Pool hypothesis reexamined’, Economic Development and Cultural Change, 48,
335–348.
Nettle, D. and Romaine, S. (2001), Vanishing Voices. Oxford: Oxford University Press.
Pool, J. (1972), ‘National development and language diversity’, in J. A. Fishman (ed.),
Advances in the Sociology of Language. The Hague: Mouton, Vol. 2, pp. 213–230.
Qizilbash, M. (2003), ‘Vague language and precise measurement: The case of poverty’,
Journal of Economic Methodology, 10, (1), 41–58.
Riedler, N. and Pons-Riedler, S. (1986), ‘An economic analysis of Canadian language
policies: A model and its implementation’, Language Problems and Language Plan-
ning, 10, 42–58.
Robinson, C. (1988), ‘Language choice: The distribution of language skills and earn-
ings in a dual language economy’, Research in Labor Economics, 9, 53–90.
Rorty, R. (1979), Philosophy and the Mirror of Nature. Princeton: Princeton University
Press.
Rossi-Landi, F. (1975), Linguistics and Economics. The Hague: Mouton.
Rubinstein, A. (2000), Economics and Language. Five Essays. Cambridge: Cambridge Uni-
versity Press.
Sally, D. (2004), ‘Can I say “bobobo” and mean “There’s no such thing as cheap talk”?’,
Journal of Economic Behavior and Organization, 245–266.
Samuels, W. J. (1990), Economics as Discourse: An Analysis of the Language of Economists.
Boston: Kluwer Academic.
Simmel, G. (1900), Philosophie des Geldes. Berlin: Duncker & Humboldt. [The Philosophy
of Money. London: Routledge and Kegan Paul, 1978].
Vaillancourt, F. (ed.) (1985), Economie et langue (Documentation du Conseil de la
language française, n 20). Quebec: Editeur Officiel du Québec.
Vaillancourt, F. (1987), ‘The costs and benefits of language policy in Québec, 1974–
1984: Some partial estimates’, in Humphrey Tonkin and Karen M. Johnson-Weiner
(eds), The Economics of Language Use (Report of the Fifth Annual Conference of the
Center for Research and Documentation on World Language Problems). New York:
Center for Research and Documentation on World Language Problems, pp. 71–93.
Vaillancourt, F. (1991), ‘The economics of language: Theory, empiricism and applica-
tion to the Asian Pacific’, Journal of Asian Pacific Communication, 2, (1), 29–67.
Walsh, J. (2006), ‘Language and socio-economic development: Towards a theoretical
framework’, Language Problems and Language Planning, 30, 127–148.
Zipf, George K. (1949), Human Behavior and the Principle of Least Effort: An Introduction
to Human Ecology. Cambridge, MA: Addison-Wesley Press.
CHAPTER
3
Linguistic Diversity and Poverty:
Many Languages and Poor People
in a Globalizing World
Suzanne Romaine
3.1 Introduction
Many of the world’s poorest people live in regions rich in linguistic diversity. By
almost any measure, e.g., the African continent (and sub-Saharan Africa in partic-
ular) is high in both linguistic diversity and poverty. Africa as a whole is home to
around 2,092 (30.3 per cent) of the world’s 6,912 languages. In 2006 more than a
third of the world’s poor lived in Africa, where mean income has been falling since
1981. Sub-Saharan Africa has the highest incidence of extreme poverty and great-
est depth of poverty; between 1981 and 2001 the number of poor people nearly
doubled from 164 to 316 million (Chen and Revallion, 2004: 141, 156). The richest
country linguistically, Papua New Guinea, is one of the world’s poorest. Its 820 lan-
guages spoken by only 0.1 per cent of the world’s population, occupying 0.4 per cent
of the world’s land area, comprise ca. 12 per cent of the world’s languages. The over-
all ratio of languages to people is only about 1 to 5,000. If this ratio were repeated in
the United States, there would be 50,000 languages spoken there.
Although facts such as these about the global distribution of linguistic diversity
and poverty are well known in the respective disciplines of linguistics and develop-
ment economics, the connections between them have not been explored, much
less understood. This chapter examines the interface between poverty and linguis-
tic diversity by looking first at the distribution of global patterns of linguistic diver-
sity, and then analyzing them in relation to various indicators of poverty and human
development. My analysis pays particular attention to indigenous peoples, who
speak around 60 per cent of the world’s languages but are among the world’s poor-
est. Section 3.7 explores the policy implications of the overlap between linguistic
diversity and poverty in today’s globalizing world.
My analysis shows how dominant discourses are embedded within a network of
systems of knowledge and power that influence how people think about develop-
ment, poverty and linguistic diversity. They construct a new world order of inter-
national relations in which we move from a kind of colonialism, whose hegemony
Linguistic Diversity and Poverty 47
is sustained primarily by direct rule and military force, to one whose power is based
on economic control of global markets. The spread of English as a global language
has played a hegemonic role in normalizing the link between English and eco-
nomic development, both at home in the major Anglophone countries as well as
abroad. In the dominant discourses linked to English, speaking English is seen as
beneficial to the world, which has freely chosen the language due to its inherent
advantages. Within this worldview, diversity has been problematized while mono-
lingualism or linguistic uniformity is ‘normalized’. Just as development and mod-
ernization have been taken to be synonymous with the introduction of western
science, technology and languages, it is still widely believed that indigenous lan-
guages are not suited for modern purposes (Romaine, 2006). Such beliefs have
prompted development strategies that result in hegemonic monocultures.
Although the fate of the world’s linguistic diversity has not figured centrally in
the world debate on globalization, poverty and inequality, addressing poverty
entails a new understanding of the critical role of language and linguistic diversity
in human development. Linguists predict that as many as 60 to 90 per cent of the
world’s 6,900 some languages may be at risk of extinction within the next 100 years
(Nettle and Romaine, 2000). While some have welcomed the loss and abandon-
ment of traditional languages and cultures as an inevitable prerequisite to modern-
ization, there can be no true development with linguistic development (Romaine,
1990). The measures most likely to preserve small languages are the very ones that
will help increase their speakers’ standard of living in a long-term, sustainable way.
3.2 Global Distribution of Linguistic Diversity
For most of the many millennia of human history people have lived in small com-
munities that are quite distinct from their neighbours. Until recently, the world was
close to linguistic equilibrium with the number of languages lost roughly equalling
the new ones created. The reason why this balance endured for so long is that
there were no massive differences among the expansionary potentials of different
peoples of the type that might cause the sustained expansion of a single, dominant
language. Over the past 10,000 years, however, a variety of events have punctured
this equilibrium forever. First, the invention and spread of agriculture, colonial-
ism, later the Industrial Revolution and today the globalization of economies and
mass media (particularly the internet) have created the global village phenome-
non. These forces have enabled a few languages – all Eurasian in origin – to spread
over the last few centuries.
Table 3.1 shows the top nine languages each with more than 100 million speak-
ers, spoken by just over 41 per cent of the world’s population. These figures include
only first language speakers. The spread of these languages would be even more
extensive if second language speakers were included.
As these large language communities expanded, others contracted. Over the
past 500 years small languages nearly everywhere have come under acute threat.
About half the known languages in the world have vanished over the past 500 years
48 Suzanne Romaine
Table 3.1 Top nine languages with over 100 million speakers (data from
Gordon, 2005).
Language Speakers % of World population
Chinese [Mandarin] 873,000,000 14.5

Spanish 322,000,000 5.4
English 309,000,000 5.1
Arabic 206,000,000 3.4
Hindi 181,000,000 3.0
Portuguese 177,000,000 2.9
Bengali 171,000,000 2.8
Russian 145,000,000 2.4
Japanese 122,000,000 2.2
Total 2,506,000,000 41.7
(Nettle and Romaine, 2000: 2). Although I have placed these pivotal events in
their chronological order in human history, they have not by any means played
themselves out to completion across the globe. For example, in Africa and else-
where the expansion of farmers is currently pushing out hunter-gatherers and
semi-nomadic pastoralists, while China and other parts of the world are just
now undergoing the industrial revolution. Before the arrival of Europeans in the
New World, there were few major empires to spread linguistic homogeneity across
the territory, and the region’s complex topography helped to maintain the distinc-
tiveness of different communities over time. Kaufman (1994: 34), for instance,
comments that ‘the linguistic diversity in the New World in 1500 was comparable
to that of Africa and Oceania of the same period, i.e., extremely diverse and at the
same time normal (italics in original)’.
The spread of large languages in modern times, however, means that there are
marked demographic disparities in the size of populations speaking the world’s
languages, as shown in Figure 3.1. If languages were equal in size, each would have
around 917,000 speakers. However, the median number of speakers for the lan-
guages of the world is only 5,000–6,000. Nearly 80 per cent of the world’s popula-
tion speaks a total of only 83 languages, each with 10 million or more speakers.
Leaving aside the world’s largest languages, however, the remaining 6,800 some
are spoken by only 20 per cent of the population. These include the smallest
languages spoken by a mere .2 per cent of the world’s population. Most, if not all,
of these may be at risk, because their speakers are among the world’s most margin-
alized peoples, who have generally been on the retreat for several hundred years.
Speakers of large languages like English and Chinese would find it difficult to
imagine the prospect of being the last speakers of their language, but the last
speakers of probably half the world’s languages are alive today. Crystal (2000: 19)
suggests that one language may vanish every two weeks. An increasing number of
LANGUAGES SPEAKERS
3,586 0.2
smallest languages spoken by
2,935 20.4%
mid-size spoken by
languages
83 79.4% of population
spoken by
Figure 3.1 Relationship between languages and speakers.
elderly people find themselves in the position of Jim Andreas, the last known per-
son who still speaks Paakanil, the native language of the Tubatulabal tribes of the
Kern River Valley of California. There are around 548 languages with fewer than
99 speakers, comprising nearly 1/10 of the world’s languages. A further 344 lan-
guages have between 10 to 99 speakers. Finally, there are 204 languages with fewer
than 10 speakers, including Ura spoken by about half a dozen elderly people on
the island of Erromango in southern Vanuatu, and Warrwa, traditionally spoken in
the Derby region of West Kimberley in Western Australia, but now spoken fluently
by only two people.
3.3 Global Distribution of Linguistic Diversity in Relation

to Poverty: Poor People in Rich Places
Over 80 per cent (N = 5,561) of the world’s languages are found in just 20 nation-
states, which include some of the richest in the world (United States, Canada and
Australia) as well as some of the poorest (Chad, Democratic Republic of the Congo
and Nigeria). Table 3.2 shows the 20 countries with the highest number of lan-
guages in relation to their ranking on the United Nation’s HDI (Human Develop-
ment Index), a comparative measure of human development widely regarded as
an international standard since the United Nations Development Program adopted
it in 1990 when it began issuing its annual report. The HDI grew out of the desire
to provide a more holistic alternative to measures based on single economic dimen-
sions such as gross domestic product or per capita income. The 2007/2008 HDI
represents a summary of average achievements of 177 countries during the year
50 Suzanne Romaine
2005 for three factors: a long and healthy life (as indicated by life expectancy),
knowledge (as measured by adult literacy rate and combined primary, secondary
and tertiary school enrolment) and standard of living (as measured by GDP, gross
domestic product per capita) (United Nations Development Program, 2007/2008:
356). Table 3.2 reveals a difference of 167 places in HDI rank between the highest
ranked country, Australia, and the lowest, Chad.
One of the most hotly contested questions in debates about the costs and bene-
fits of globalization is whether economic growth around the world is good for the
poor. Proponents of globalization have emphasized as one of its triumphs a nearly
continuous rise in standard of living and an overall reduction of poverty. Using a
rather frugal yardstick of number of people living on US$1 a day (the United Nation’s
criterion for absolute poverty), there were 1.1 billion poor people in 2001, which
represents a decline of 400 million over the previous two decades, or a near halving
of the 1981 poverty rate of 40 per cent (Chen and Revallion, 2004: 141,162).
Table 3.2 Top 20 countries with highest number of languages and HDI rank (Data com-
piled from Gordon 2005 and Human Development Report, 2007/2008: 356.).
Country N languages % of world’s N speakers HDI rank

languages
Papua New Guinea 820 11.86 3,665,383 145

Indonesia 742 10.73 218,607,876 107
Nigeria 516 7.47 90,026,548 158
India 427 6.18 943,283,395 128
United States 311 4.5 262,740,392 12
Mexico 297 4.3 92,701,909 52
Cameroon 280 4.05 9,638,055 144
Australia 275 3.98 16,781,815 3
China 241 3.49 1,226,606,396 81
Democratic Republic 216 3.12 37,945,510 168
of the Congo
Brazil 200 2.89 165,836,223 70
Philippines 180 2.6 70,556,507 90
Malaysia 147 2.13 15,676,135 63
Canada 145 2.1 27,388,756 4
Sudan 134 1.94 23,492,482 147
Chad 133 1.92 5,909,406 170
Russia 129 1.87 139,356,854 67
Tanzania 128 1.85 25,837,668 159
Nepal 125 1.81 22,983,905 142
Vanuatu 115 1.66 119,759 120
TOTAL 5561 80.45 3,399,154,974
Nevertheless, not everyone agrees that globalization is reducing poverty and

inequality (Deaton, 2002, 2006). Immense growth in China and India has resulted
in a reduction of the number of poorest people worldwide, but evidence that a
rising tide does not lift all boats can be found in the fact that since the mid-1970s
HDI has been progressively increasing in almost all regions with one major excep-
tion, sub-Saharan Africa, which has stagnated since the early 1990s. The bottom
24 countries are all African. The bottom five countries in terms of HDI in Table 3.2
are also all African (i.e. Chad, Democratic Republic of the Congo, Tanzania,
Nigeria and Sudan). Moreover, four of the five countries in Table 3.2 experienced
HDI reversal over the period from 1990 to 2003, including Cameroon, Democratic
Republic of Congo, Russia and Tanzania (United Nations Development Program,
2005: 21, Table 1.1).
The United Nations Development Program (2006: 273) identified as one of the
central human development challenges in the decades ahead the need to ‘dimin-
ish the tolerance for extreme inequalities that have characterized globalization
since the early 1990s and to ensure that the rising tide of prosperity extends oppor-
tunities for the many, and not just the privileged few’. The poorest one-fifth of
humanity living on only $1 a day receives only 1.5 per cent of global income. The
situation is particularly acute in sub-Saharan African, where one in two people is in
this poorest 20 per cent; and this region’s share of those in the bottom 20 per cent
has more than doubled since 1980 to 36 per cent of the total. At the other end of
the spectrum, nine in ten people living in the high-income countries of the OECD
(Organization for Economic Cooperation and Development) are in the top 20 per
cent of global income distribution. The assets of the world’s top three billionaires
amount to more than the combined gross national products of all the least devel-
oped countries and their 600 million people. Such disparities indicate that ‘wealth
accumulation at the top of the global income distribution has been more impres-
sive than poverty reduction at the bottom’ (United Nations Development Program,
2006: 269). Thus, it will take much more growth in Africa than elsewhere to have
the same impact on poverty levels.
Poverty is not just geographically specific, it is socially specific. Closer exami-
nation of individual countries reveals further deep socio-economic disparities.
Although the poverty line in the United States is more than ten times as much as
the absolute ‘extreme poverty’ line of $1 a person a day used by the World Bank
and the United Nations, poverty is nevertheless relative. A nation of poor people
the size of Canada’s population lives in the United States, and the number below
the so-called poverty line is increasing. In 2005 the number of people below the
poverty line (defined as $14,680 annual income for a family of three) hit 37 million.
According to this metric about 8 per cent of white Americans, 22 per cent of
Hispanics, and 25 per cent of African-Americans are poor (Alter, 2005: 44).
The world saw a dramatic illustration of the poor ‘nation within’ during the
Katrina disaster and its aftermath. Photos from New Orleans and the surrounding
Gulf coast communities looked more like scenes from the 2004 Indian Ocean
tsunami. Those most devastated and left to fend for themselves were the poor, the
elderly, the sick, most of them African-Americans. As the world wondered how this
52 Suzanne Romaine
could happen in the richest country in the world, President Bush, in his address to
the nation on 15 September 2005, acknowledged ‘some deep, persistent poverty’
that ‘has roots in a history of racial discrimination, which cut off generations
from the opportunity of America’. He told Americans that ‘we have a duty to con-
front this poverty with bold action . . . to ‘rise above the legacy of inequality’ (Bush,
2005).
This example illustrates that the way we in which we calculate, benchmark and
monitor improvements (or declines) across countries and regions can hide as much
as it reveals. Disparities are disguised in HDI, which looks at country averages
rather than at the distribution of human development within countries. Table 3.3
looks at the difference between the poorest 20 per cent and the national average
in 2005 HDI rank for selected linguistically diverse countries from Table 3.2.
The disparity between the HDI ranking of the poorest 20 per cent in the United
States and the country as a whole is 21. This means that the poorest 20 per cent in
the United States compare to citizens of the Czech Republic. Since 1979 income
inequality has been rising in the United States, which now has the greatest level of
income inequality among the 20 highest ranking countries, and is on a par with
Ghana and Cambodia. There is a 33-year gap between those with the highest and
lowest life expectancies. Native American men have the lowest life expectancy and
live shorter lives than men in many third world countries, such as Bangladesh
(Murray et al., 2006).
The relativity of poverty defined in economic terms means that even the bottom
10 per cent or 20 per cent in a very rich country such as the United States are pros-
perous when compared to the best off in a country that is overall much poorer
such as Brazil, in 63rd place in the global HDI ranking. The poorest 20 per cent of
the population in Brazil rank 115, which is 52 places lower than the average for the
country. This means that the situation of the Brazilian poorest 20 per cent is com-
parable to that in countries such as Guatemala, Honduras and Mongolia. There
is an even greater difference among the poorest 20 per cent in Mexico, who are
comparable to Vietnam as a whole (United Nations Development Program, 2005:
333, Table 2). The Russian poorest 20 per cent are on a par with the Dominican
Republic, while the Chinese poorest are on a par with the Republic of Moldova.
Table 3.3 Difference between poorest 20 per cent and national average
in 2005 HDI rank for selected linguistically diverse countries (data from
the United Nations Development Program, 2005: 333, table 2).
Country HDI rank Poorest 20% Difference
Brazil 63 115 52
China 85 115 30
Mexico 53 108 55
Russia 62 95 33
United States 10 31 21
3.4 The Poorest of the Poor: Poverty and Linguistic Diversity

among Indigenous Peoples
The situation of the world’s 370 some million indigenous and tribal peoples is typi-
cally not reflected in statistics for a variety of reasons or is hidden by national aver-
ages such as HDI. Indigenous peoples are often the most marginalized peoples in
society, without adequate access to education, health care and water. They are fre-
quently not recognized by the governments of the states in which they reside and
thus are deprived of the right to participate in or direct their own sustainable
human development.
A case in point are the 410,000 Aboriginal Australians encapsulated within a
country with the highest HDI ranking in Table 3.2. The position of Aboriginal
Australians in terms of health and poverty indicators is usually obscured in official
statistics because they are a very small proportion, less than 3 per cent, of the coun-
try’s total population. Although the incomes of Aboriginal Australians are in excess
of US$1 per day, the substantial income gap between indigenous Australians and
the rest of the Australian population goes hand in hand with a variety of other
indicators incorporated in HDI such as reduced life expectancy, low schooling
rates and low incomes (Taylor, 2007). The life expectancy for Aboriginal Austral-
ians is 17 years lower than for non-Aboriginals, which puts them on a par with peo-
ple in other ‘low income countries’. This means that life expectancy for Aboriginal
people is similar to that of people in Papua New Guinea, a former Australian colony,
and considerably lower than in Nigeria, Nepal, India and Bangladesh. Education
levels for Aboriginal people are also lower than for non-Aboriginals; Aboriginal
Australians are only half as likely to have completed 12 years of school, i.e. 18 per
cent compared with 41 per cent.
Likewise in Canada, the country in Table 3.2 with the second highest HDI rank-
ing, the differences between native people (i.e. First Nations, Métis and Inuit) and
other Canadians are substantial. On average, they rank 48th in terms of HDI, on a
par with Latvia. In 2000 the life expectancy for the registered Indian population
was 68.9 years for men and 76.6 years for women, a difference of 7.4 and 5.2 years,
respectively, from the figures for the non-Indian population (Health Canada, 2005).
Registered Indians also rank lower than the general Canadian population on all
indicators of educational attainment, including completion rates for secondary
school and postsecondary entrance and completion rates (Census of Canada,
2001). Most native people in Canada also live at or below the poverty line.
Far more disturbing, however, are differences in potential years of life lost
to suicide and unintentional injuries. In 2000, the suicide rate among First Nations
was nearly three times that of the general Canadian population. Suicide and self-
injury were the leading causes of death among youth and adults up to 44 years of
age (Health Canada, 2005). High suicide rates are found among many indigenous
groups around the world, especially among young people, but in Canada, First
Nations and other Aboriginal youth take their own lives at higher rates than those
found for any culturally identifiable group in the world (Kirmayer, 1994). The suicide
rate for young men is 8.3 times higher and for young women 20 times higher than
54 Suzanne Romaine
the already elevated rate among Aboriginal people as a whole (Chandler et al.,
2003: 32).
Similarly, in the United States, suicide rates among American Indian/Alaskan
Native adolescents and young adults between the ages of 15 and 34 are 1.9 times
higher (i.e. 21.4 per 100,000) than the national average for that age group (11.5 per
100,000). In this age group suicide is the second leading cause of death (Centers
for Disease Control, 2007). Suicide rates remained unchanged for native Americans
between 1989 and 1998. The largest number of youth suicides occurs among Alaska
natives, whose rate is eight times higher than US national rates (Patel et al., 2005).
In Australia the rate of suicide and suicidal behaviours started going up in 1970s,
especially for young males. In 2005, suicide accounted for 4.3 per cent of all
Aboriginal deaths compared with 1.6 per cent of deaths for other Australians.
The United Nations Development Program (2006: 273) concluded that ‘the
increasingly visible divides that separate the haves and the have-nots in rich coun-
tries have become a focal point for discontent’. If such stark disparities exist
between the indigenous and dominant populations in the richest countries, what
about some of the poorest ones in Table 3.2? What is happening in Mexico is
instructive, as it has the largest indigenous population in Latin America in abso-
lute terms; 11 per cent of the population belongs to around 62 language groups.
Although Mexico has made great progress in poverty reduction since the 1990s,
the Human Development Report (2005) shows that 92.4 per cent of the population
with the lowest HDI is indigenous. The lowest ranking of the 50 municipalities
compares to Malawi, whose HDI of 165 puts it within the bottom 20 countries.
Hence, in Mexico too poverty affects indigenous people disproportionately and
indigenous people benefit less from poverty reduction efforts. The same is true for
some of India’s scheduled tribes, whose HDI is comparable to sub-Saharan African
countries (Tauli-Corpuz, 2006).
In Africa, home to 960 million people and six of the world’s twenty most linguis-
tically diverse countries, the situation of indigenous people is nothing short of
dire. Indigenous health is systematically worse than among the non-indigenous
population, particularly where loss of land and customary resource bases has
rendered people unable to maintain their traditional livelihoods and cultural
practices, including their languages. In the case of the 300,000 to 500,000 Pygmy
people, who live in ten central African countries, those who are able to maintain
their traditional forest-based life have better health than those who have lost access
to the forest through logging and farming. Forests are a vital component of a
Pygmy sense of physical and spiritual well-being. Pygmy communities living out-
side the forest in fixed settlements are unable to meet their food needs and experi-
ence higher rates of infectious diseases such as malaria, tuberculosis, HIV/AIDS
and parasites. In the absence of traditional cultural practices reducing social ten-
sions, domestic violence against women and alcohol abuse have increased (Ohenjo
et al., 2006: 1939–1941).
The San are widely recognized as the most impoverished, disempowered and
stigmatized ethnic group in Southern Africa. The HDI for 32,000 some San living
in Namibia, where the second largest population of San resides, is not only the
lowest, but they are also the only group whose HDI fell between 1996 and 1998.
Health and welfare among the San in resettlement areas have declined; alcohol
consumption and violence against women have increased. Ohenjo et al. (2006:
1943) regard these trends ultimately as ‘a problem of poverty stemming from the
loss of land and livelihoods without a viable alternative’.
The largest population of San numbering some 47,000 live in Botswana, where
they comprise about 4 per cent of the population of Botswana. Despite the fact that
Botswana has the fourth highest per capita income in Africa, the poorest 10 per
cent of the population consists largely of San and related minority communities,
who receive only 0.7 per cent of the nation’s income (Ohenjo et al., 2006: 1942).
The National Constitution makes no reference to the San among the main eight
tribes of the country. The government has justified its relocation of San people
from the Central Kalahari Game Reserve by claiming that the San deplete the nat-
ural resources of the reserve and that providing services to them on the reserve is
too expensive. Since their displacement, the San have been unable to adapt to
their new surroundings, where they have no means of subsistence, and are increas-
ingly dependent on the government for food relief and cash-for-work programs.
3.5 Reconciling Development and Linguistic Diversity
The closer development comes to the poorest, most marginalized peoples in the
world, the more likely it is that they will speak a different language from their neigh-
bours. Consider Cameroon, one of the most linguistically diverse and economically
undeveloped African countries. For the Nugunu people in Ombessa, the idea of
development, like French, the language in which it is most often offered and deliv-
ered, is seen as belonging to official spheres of activity (Robinson, 1996: 167). Like
French, development is remote and not theirs. Villagers with little status rarely
have any interaction with development agencies or their agents; only 4 per cent
(mostly men) had contact personally and in organized meetings. Development
thus reaches local male elites more readily than the ordinary villager. These same
local elites also have the highest level of schooling and greatest competence
in French (Robinson, 1996: 212–213). The educational system relies exclusively
on French. Thus, development encounters exacerbate a self- perpetuating cycle
involving language, development and education.
Just as children cannot benefit from schooling in a language they do not under-
stand, the rural poor cannot benefit from development assistance rendered in a
language they do not understand. In most parts of the world schooling is still virtu-
ally synonymous with learning a second language, particularly in Africa, where few
African languages are used in government or education. Robinson (1996: 219–223)
shows how language use is intimately connected with the exercise and distribution
of power and control in meetings held by development agents and other officials.
Where French is used, the agenda and discussion remains outside the control of
villagers. One official told Robinson that communication problems in the village
would be solved when ‘everybody had learnt enough French’ (Robinson, 1996: 221).
The cause of the problem is offered as the solution.
56 Suzanne Romaine
Use of local languages is inseparable from participatory development. By rely-

ing on French, development activities and resources reach those who are relatively
well educated and whose access to resources is already higher, all the while margin-
alizing the less advantaged still further. Despite the widespread recognition of
women’s lower levels of literacy, access to schooling, etc., development does not
address the fact that speaking a minority language compounds the marginaliza-
tion of being a woman (Robinson, 1996: 216).
The continuation of educational policies favouring international languages at
the expense of local ones is part of the development fiasco (Romaine, 1990). Yet
few have asked the rural poor what their conceptions and understanding of devel-
opment are. For this to occur, it would, of course, be essential to conduct on the
ground surveys in local languages before planning intervention projects. This is
seldom done, despite increasing calls for development to come from within com-
munities, and to be based on local indigenous knowledge, so that it is responsive
to self-perceived needs, is culturally relevant and sustainable.
Although the conventional wisdom on which most development theories
are premised often assumes that people and places are poor because they lack
resources, poverty is clearly a complex phenomenon. Cause and effect are often
confused in arguments that poverty is caused by lack of infrastructure, services
and employment opportunities in the rural communities where most indigenous
peoples live. Thus, prevailing policies have generally entailed assimilation and
integration into the dominant society’s modes of production and employment,
often requiring migration or forced relocation to urban areas. Such strategies are
motivated by misperception and stereotyping of indigenous peoples, their lifeways
(particularly hunter-gatherers) and languages as primitive, backward and an obsta-
cle to development. President Festus Mogae of Botswana, for instance, asked, ‘How
can you have a stone-age creature continuing to exist in the time of computers? If
the Bushmen want to survive, they must change, otherwise, like the dodo they will
perish’) (cited in Ohenjo et al., 2006: 1942).
However, the experience of the San, Pygmies and most other indigenous
peoples shows that if people are deprived of traditional means of subsistence, but
unprepared for employment in other than the lowest and poorest paid jobs,
a declining quality of life will result from development projects and assimilation
policies (Wilmer, 1993:133). Taylor (2007: 16), for instance, draws attention to
what he sees as
a clear contradiction between the desire of many Indigenous people to live

in remote areas in small dispersed communities on traditional lands, and the
general thrust of government policy that is intent on securing Indigenous par-
ticipation in the mainstream urban economy as the core means to enhance
well-being.
Many indigenous people define themselves in terms of their close relationships

to land and community, where a good life is associated with maintaining tradi-
tional hunting, gathering and herding practices. Many of the Arctic’s residents,
e.g., would not want to exchange this way of life for the lifestyles of residents of
southern metropolises, even though such a life may offer higher standards of liv-
ing in material terms (AHDR, 2004: 16–17). For many, well-being is to be found in
a way of life that minimizes the need for the sorts of material goods and services
included in calculations of GDP per capita. Mainstream measures of well-being
based primarily on economic indicators are therefore often at odds with indige-
nous perceptions and goals. Income-based measures of poverty like GDP per
capita often fail to capture the health of subsistence systems or mixed economies.
In focusing only on gaps between indigenous and mainstream majority popula-
tions in terms of economic measures, other factors such as discrimination and the
significance of indigenous perceptions and priorities are often ignored and/or
discounted. Wealthy nations continue to set the agenda for development, define
its scope, and monitor its progress in terms of quantitative supposedly universal
indicators of development. This worldview pervades the education sector as well,
where the primary aim of education is seen as a means of economic development,
driven by the need for a highly trained workforce.
As the world’s only truly global language, English is paradoxically positioned as
both the key to and an obstacle to this development. Ironically, the United Nations
Development Program (2006: 263) observes that
enthusiasts who emphasize the positive aspects of globalization sometimes get

carried away. They increasingly use the language of the global village to describe
the new order. But when viewed through the lens of human development the
global village appears deeply divided between the streets of the haves and those
of the havenots.
In its guise as the world’s most important language of the post-industrial global
village, English is seen as the epitome of a modern language, the road to develop-
ment, economic prosperity and freedom. Examples are readily found of this
discourse, such as an editorial from The Wall Street Journal (30 December 2004)
entitled ‘Se habla ingles’, reporting on Chile’s educational program to ensure that
all students graduating from high school are fluent ‘in the globe’s international
language’. Chile’s minister of education was reported as saying ‘we know our lives
are linked more and more to an international presence, and if you can’t speak
English, you can’t sell and you can’t learn.’ The editorial goes on to say that
although ‘nationalists’ are calling the plan ‘an affront to culture’,
market-friendly Chile seems to grasp the finer point: Speaking English will make
Chile, already a world-class exporter, an even more daunting competitor in
world trade. The result will be a richer Chile better able to define and defend
Chilean culture. Poverty and isolation are not cultural protection.
Chile’s commitment to making English ‘part of a standard education is a pledge

that no child will be left behind’. In this parlance English ceases to be associated
with a particular group to become ‘the globe’s international language’. Those who
58 Suzanne Romaine
oppose an increasing role for English are marked as ‘nationalist’, while the claims
made for English are seen to be neutral, unmarked and self-evidently in the best
interests of Chile as a country and of its school children. Note how the activities of
selling and even learning in general are viewed as dependent on English. Although
Spanish is not specifically mentioned, except in the editorial’s facetious title, there
is an implied linkage between Spanish, poverty and isolation on the one hand, and
English, global trade and a ‘richer Chile’ on the other. No mention is made of
Chile’s 600,000 Mapuche people, who account for about 4 per cent of the popula-
tion, but whose HDI is lower than that of other Chileans.
Another example of the language of the global village being used to describe
the new order comes from Martínez-Fernández (2006: 20), who explained that
one reason why ‘a person whose native language is not English can adopt the
English language as a means of communication’ is ‘the need to use a more precise
language with a richer vocabulary. (English has about 900,000 words’, while French,
e.g., has fewer than 100,000.)’ Here spurious word counts, for which no evidence
is cited, are used to lend support to a Darwinian conceptualization of linguistic
evolution as a natural process of survival of the fittest. If size matters, the bigger the
better. These ideas fit easily into a larger narrative in which indigenous languages
and cultures are dismissed as primitive and backward-looking, which is then used to
justify their replacement by western languages and cultures as prerequisites to mod-
ernization and progress (Romaine, 2008a).
3.6 Policy Implications of Poverty and Linguistic Diversity
Impediments to participation in and right to exercise control over development

processes affecting peoples such as the San are increasingly being recognized
internationally as human rights violations. Former United Nations High Commis-
sioner for Human Rights, Mary Robinson made explicit the linkage between pov-
erty and the denial of human rights when she observed: ‘I am often asked what is
the most serious form of human rights violations in the world today, and my reply
is consistent: extreme poverty.’ (United Nations Development Program, 2003: iv).
As defined by all the major international treaties and other legal instruments,
human rights are inalienable entitlements inherent to the person and belong
equally to all human beings. As embodied above all in the 1948 Universal Declara-
tion of Human Rights, these consist of both civil and political rights as well as eco-
nomic, social and cultural rights. The inalienability of our common entitlement
follows from the fact that we are all human, and therefore all the same, but para-
doxically the need to guarantee them in law arises from the fact that we are diversely
different (Romaine, 2008c).
The interdependence of these two domains of human rights emerges nowhere
more clearly than in the global debate about poverty reduction in the context of
the eight Millennium Development Goals (MDGs) adopted by 147 nations of the
UN General Assembly at the United Nations Millennium Summit in 2000. Despite
the United Nations Development Program’s (2000: 8) characterization of poverty
eradication as ‘a central challenge for human rights in the 21st century’, and its
forceful reiteration of the maxim that ‘poverty is a denial of human rights’ (United
Nations Development Program, 2003: iv), these principles have taken backstage
to economic growth in what has become the largest and arguably most ambitious
initiative on the international development agenda. Indeed, the United Nations
Development Program (2006: v) regards the MDGS as a normative framework for
human development.
However, the MDGs, like most other large development undertakings, were not
initiated by poor countries, but were prompted primarily by the United States,
Europe and Japan. MDGs assume that poverty will be reduced by growth in GDP
through top down infusion of aid. Targeting funding to economically disadvan-
taged areas and populations does not directly address the socio-economic and cul-
tural processes and policies that lead to inequalities and poverty in the first place.
Despite the rights-conscious approach articulated in the Human Development
Reports, which stress that human rights are not ‘optional extras’, but rather ‘bind-
ing obligations that reflect universal values and entail responsibilities on the part
of governments’ (United Nations Development Program, 2006: 4), the discourse
of free market economics is pre-empting and displacing that of human rights
(Romaine, 2008b).
Such is the power exercised by the dominant discourse within international
development circles that it would be hard to find anyone who disagrees with an
agenda dedicated to economic growth. Who would not want to reduce poverty by
half by 2015? Yet paradoxically, the goal of halving the number of people living on
less than $1 a day has been criticized by some as too ambitious and by others as not
ambitious enough. Setting the benchmark at only halving poverty rather than
eliminating it altogether leaves open the possibility that minorities will constitute
the majority of those persons still living in poverty in 2015.
At its Fourth and Fifth Sessions, the United Nations Permanent Forum on Indig-
enous Issues (UNPFII) expressed concern that indigenous issues were often absent
from the MDGs. Endorsing a human rights-based approach that takes into account
poverty indicators based on indigenous peoples’ own perception, members placed
high priority on the right of indigenous peoples to sustainable development and
called for relations between development agencies and indigenous peoples to be
direct rather than mediated through institutions of the dominant society. They
recommended promotion of multicultural policies, affirmative action and special
measures for indigenous peoples along with safeguards ensuring their participa-
tion in planning, implementation and monitoring of all projects and policies
(United Nations Permanent Forum on Indigenous Issues, 2005, 2006).
Acknowledging that ‘integrating human rights in poverty reduction strategies
does not so much changes [sic] “what” is to be done as to “how” and “why” activi-
ties are undertaken’ (United Nations Development Program, 2003: iv), it is worth
weighing up the economic and human costs of continuing business as usual. While
some have welcomed the loss and abandonment of traditional languages and
cultures as part of an inevitable march to economic progress, discontinuities in
transmission of culture and language are frequently accompanied by large human
60 Suzanne Romaine
and social costs manifested in poverty, poor health, drug and alcohol abuse, family
violence and suicide. Although scholars and community members themselves have
long argued for a specific link between indigenous language loss and community-
level measures of health and well-being, Hallett et al. (2007: 394) have provided
empirical evidence to support the claim that ‘the generic association between
cultural collapse and the rise of public health problems is so uniform and so excep-
tionless as to be beyond serious doubt.’
First Nations bands in British Columbia where fewer than 50 per cent reported
conversational language knowledge had more than six times (96.59 per 100,000)
the number of suicides as communities with higher levels of language knowledge.
Among the latter the suicide rate was 13.00 per 100,000, well below the provincial
average for both Aboriginal and non-Aboriginal youth (Hallett et al., 2007: 396).
In fact, reported language knowledge proved to have predictive power over and
above that of six other cultural continuity factors including community control
over the delivery of health, education, child protection and policing services,
achievement of a degree of self-governance, secure access to traditional lands, and
construction of facilities for preserving cultural artefacts and traditions (Chandler
et al., 2003, Lalonde, 2006).
When communities are successful in promoting their cultural heritage, they are
better positioned to claim ownership of their past and future. The positive effects
reverberate in a variety of measures relating to youth health and welfare: suicide
rates fall, along with rates for intentional injuries. Fewer children get taken into
care and school completion rates rise. Now we have proof that language mainte-
nance matters too. Most importantly, these studies provide strong empirical evi-
dence against the view that assimilation to dominant cultures in the interests of
modernity is harmless or even beneficial to individuals and communities in the
ways suggested by prevailing policies.
Although this research was conducted among 152 of the 195 First Nations bands
in British Columbia, the Canadian province with the largest number of endan-
gered languages as well as the smallest language populations, the findings are
clearly relevant not only for other native communities in Canada but for indige-
nous peoples everywhere experiencing language shift. According to the 2001
Canadian Census, overall only 15 per cent of the country’s Aboriginal children
learn an indigenous mother tongue, and fewer still are spoken to in such a lan-
guage at home. Global estimates for language loss range from 50 to 90 per cent of
the world’s 6,900 some languages, with the heaviest burden falling on indigenous
peoples who speak at least 60 per cent of them (Nettle and Romaine, 2000).
Lalonde (2006) believes that the success achieved by some First Nations com-
munities in helping their young people has clear implications for policy makers
and service providers. Because forms of indigenous knowledge have proven their
worth in First Nations communities, he contends that the best chances for success
lie in efforts to reassert cultural sovereignty and to expand the indigenous knowl-
edge base. Writing from the perspective of indigenous knowledge, Burgess (1999: 23)
reaches a similar conclusion when he writes that the real goal ‘is to focus resources
on areas that sustain healthy communities and the maintenance of traditional
activities . . . as part of a process which is intimately linked to the quest for self-
determination elements of indigenous peoples’. Likewise, in his comparative
survey of indigenous peoples and poverty in Australia, Canada, New Zealand and
the United States, Cornell (2006) contends that governments’ refusals to come
to grips with indigenous demands for self-determination cripple efforts to over-
come indigenous poverty. Self-determination and self-government are profoundly
connected, and public policy has to take this into account.
Despite this convergence of findings on the importance of securing recognition
of indigenous rights, recent international developments continue to undermine
these efforts. For instance, in 2006 the Canadian government withdrew funding
promised for the support of Aboriginal languages, and in 2007 Canada, the United
States, Australia and New Zealand vetoed the UN declaration of rights for indige-
nous peoples.
3.7 Conclusion
This chapter has shown that the distribution of both linguistic diversity and poverty
is strikingly uneven, geographically, demographically as well as socio-economically.
The neoliberal ideology underpinning the MDG approach to poverty reduction
through prioritizing economic growth has propelled human rights and develop-
ment down increasingly separate paths. The rhetoric portraying western models of
government, development and education as the only viable ones for contempo-
rary societies continues the colonizing agenda of the past. Rejecting that coloniza-
tion entails rejecting the discourse as well. The nation-state remains the most
critical unit of analysis because it is the policies pursued within national bounda-
ries that give some languages (and their speakers) the status of majority and others
that of minority language. The fact that the only institutions with authority to regu-
late language policies exist within the political bodies of individual states makes
planning and policy at the supranational level difficult. The effectiveness of any
initiatives on the supranational level can always be undermined by individual states
unless there is some way of guaranteeing the implementation of language-related
measures on a supranational level (Romaine, 2007).
It is time for a reconceptualization of the MDGS and a new understanding of
poverty and development. Maintaining the world’s languages goes hand in hand
with achieving and maintaining greater self-determination as part of a larger strat-
egy of cultural survival and a much larger healing journey.
References
Alter, J. (2005), ‘Poverty in America: The truths that Katrina laid bare’. Newsweek,
CXLVI, (12), 42–49. 19 September 2005.
AHDR (Arctic Human Development Report) (2004), Akureyri: Stefansson Arctic Institute.
Burgess, P. (1999), Traditional Knowledge. A report prepared for the Arctic Council
Indigenous Peoples’ Secretariat, Copenhagen.
62 Suzanne Romaine
Bush, G. W. (2005), Transcript of address to the nation on recovery from Katrina.

Electronic document, accessed 19 September 2005. http://www.whitehouse.gov/
news/releases/2005/09/20050915-8.html
Census of Canada 2001. Ottawa: Statistics Canada.
Centers for Disease Control (2007), Suicide. Electronic document, accessed 24 November
2007. http://www.cdc.gov/ncipc/dvp/Suicide/SuicideDataSheet.pdf
Chandler, M. J., Lalonde, C. E., Sokol, B. W. and Hallett, D. (2003), Personal Persistence,
Identity Development, and Suicide: A Study of Native and Non-native North American
Adolescents. Monographs of the Society for Research in Child Development. Serial
No. 273. Volume 68. No. 2.
Chen, S. and Ravallion, M. (2004), ‘How have the world’s poorest fared since the early
1980’s?’, The World Bank Research Observer, 19,(2),141–169.
Cornell, S. (2006), ‘Indigenous peoples, poverty and self-determination in Australia,
New Zealand, Canada and the United States’, in R. Eversole, J-A. McNeish, and
A. D. Cimadamore (eds), Indigenous Peoples and Poverty: An International Perspective.
London: Zed Books in association with CROP International Studies in Poverty
Research, chapter 11.
Crystal, D. (2000), Language Death. Cambridge: Cambridge University Press.
Deaton, A. (2002), ‘Is world poverty falling?’, Finance and Development, 39,(2),4–7.
Deaton, A. (2006), ‘Measuring poverty’, in A. Banerjee, R. Benabou, and D. Mookerjee
(eds), Understanding Poverty. Oxford: Oxford University Press, pp. 3–16.
Gordon, R. G., Jr. (ed.) (2005), Ethnologue: Languages of the World. Fifteenth edition.
Dallas: SIL International. Online version: http://www.ethnologue.com/, accessed
19 November 2007.
Hallett, D., Chandler, M. J., and Lalonde, C. E. (2007), ‘Aboriginal language knowl-
edge and youth suicide’, Cognitive Development, 22,(3), 392–399.
Health Canada (2005), A Statistical Profile on the Health of First Nations in Canada for the
Year 2000. Ottawa.
Indian and Northern Affairs Canada (2001), Basic Developmental data 2001. Ottawa.
Kaufman, T. (1994), ‘The native languages of Meso-America’. in C. Moseley and
R.E. Asher (eds), Atlas of the World’s Languages. London: Routledge, pp. 34-41.
Kirmayer, L. J. (1994), ‘Suicide among Canadian aboriginal people’, Transcultural
Psychiatric Research Review, 31, 3–57.
Lalonde, C. E. (2006), ‘Identity formation and cultural resilience in Aboriginal com-
munities’, in R. J. Flynn, P. Dudding and J. G. Barber (eds), Promoting Resilient Devel-
opment in Young People Receiving Care: International Perspectives on Theory, Research,
Practice and Policy. Ottawa: University of Ottawa Press, pp. 52–71.
Martínez-Fernández, L. (2006), ‘Just like us? Not likely’, Chronicle of Higher Education
53, (16), B20.
Murray, C. J. L., Kulkarni, S. C., Michaud, C., Tomijima, N., Bulzacchelli, M. et al.
(2006), ‘Eight Americas: Investigating mortality disparities across races, counties,
and race-counties in the United States’, PLoS Med, 3,(9), e260. Electronic docu-
ment, accessed 21 November 2007. DOI: 10.1371/journal.pmed.0030260.
Nettle, D. and Romaine, S. (2000), Vanishing Voices. The Extinction of the World’s Languages.
Oxford: Oxford University Press.
Ohenjo, N., Willis, R., Jackson, D., Nettleton, C., Good, K. and Mugarura, B. (2006),
‘Health of indigenous people in Africa’, The Lancet, 367, (9526), 1937–1946.
Patel, R. N., Wallace, L. J. D. and Paulozzi, L. (2005), Atlas of Injury Mortality among
American Indian and Alaska Native Children and Youth, 1989–1998. US Department of
Health and Human Services.
Robinson, C. D. W. (1996), Language Use in Rural Development. An African Perspective.
Berlin: Mouton de Gruyter.
Romaine, S. (1990), Language, Education and Development: Urban and Rural Tok Pisin in
Papua New Guinea. Oxford: Oxford University Press.
Romaine, S. (2006), ‘Planning for the survival of linguistic diversity’, Language Policy,
5(2), 443–475.
Romaine, S. (2007), ‘The impact of language policy on endangered languages’, in
M. Koenig and P. De Guchteneire (eds), Democracy and Human Rights in Multicultural
Societies. Aldershot: Ashgate/UNESCO. Chapter 10, pp. 217–236.
Romaine, S. (2008a), ‘Biodiversity, linguistic diversity, and poverty – some global patterns
and missing links’, in W. Harbert, S. McConnell-Ginet, A. J. Miller and J. Whitman
(eds), Language and Poverty. Clevedon: Multilingual Matters, pp. 127–146.
Romaine, S. (2008b), ‘Language rights, human development and linguistic diversity in
a globalizing world’, in P. Van Sterkenburg (ed.), Unity and Diversity of Languages.
Amsterdam: John Benjamins, 85–96.
Romaine, S. (2008c), ‘Linguistic diversity, sustainability, and the future of the past’, in
K. King, N. Schilling-Estes, L. Fogle, J. Lou, and B. Soukup, B. (eds), Sustaining
Linguistic Diversity. Endangered and Minority Languages and Varieties. Washington, DC:
Georgetown University Press, pp. 7–21.
Tauli-Corpuz, V. (2006), Address to the opening of the fifth session of the Permanent
Forum on Indigenous Issues. New York. 15 May. Electronic document, accessed
20 November 2007. http://www.un.org/esa/socdev/unpfii/en/session_fifth.html
Taylor, J. (2007), ‘Indigenous peoples and indicators of well-being: Australian perspec-
tives on United Nations global frameworks’, Social Indicators Research, 82, 1–16.
United Nations Development Program (2000), Human Development Report. Human
Rights and Human Development. New York: United Nations Development Program.
United Nations Development Program (2003), Poverty Reduction and Human Rights:
A Practice Note, June 2003. New York: United Nations Development Program.
United Nations Development Program (2005), Human Development Report. International
Cooperation at a Crossroads. Aid, Trade and Security in an Unequal World. New York:
United Nations Development Program.
United Nations Development Program (2006), Human Development Report. Beyond
Scarcity: Power, Poverty and the Global Water Crisis. New York: United Nations Develop-
ment Program.
United Nations Development Program (2007/2008), Human Development Report. Fighting
Climate Change: Human Solidarity in a Divided World. New York: United Nations Devel-
opment Program.
United Nations Permanent Forum on Indigenous Issues (2005), Millennium Develop-
ment Goals and Indigenous Peoples with a Focus on Goal 1 to Eradicate Poverty and Extreme
Hunger, and Goal 2 to Achieve Universal Primary Education. Electronic document,
64 Suzanne Romaine
accessed 20 September 2005. http://www.un.org/esa/socdev/unpfii/en/session_

fourth.html
United Nations Permanent Forum on Indigenous Issues (2006), Millennium Development
Goals and Indigenous peoples: Redefining the Millennium Development Goals. Electronic
document, accessed 20 November 2007. http://www.un.org/esa/socdev/unpfii/
en/session_fifth.html
Wall Street Journal (2004), ‘Se habla ingles’, 30 December.
Wilmer, F. (1993), The Indigenous Voice in World Politics. Newbury Park: Sage.
CHAPTER
4
Religious Language Management
Bernard Spolsky
4.1 Introduction
One of the most widely noted international language management actions of

recent years was the decision of Vatican II to allow conducting the mass in the
vernacular rather than in the traditional Latin. The fact that Arabic is so widely
spoken today is accounted for by the insistence of Islam that all religious services
be conducted in it. Hebrew was kept alive for nearly two millennia after people
stopped speaking it by continuing use as language of prayer and religious learning.
In much of Africa, the current sociolinguistic situation owes much to arbitrary deci-
sions by missionaries on which local dialects to standardize for Bible translation. All
of these point to the central role that religion and religious institutions have played
in language management. However, probably as a result of the secularization of
western scholarship, the field has not received the attention it deserves.
Religious institutions have been the focus of language conflict. One of the issues
of dispute between Reform and Orthodox Judaism in the nineteenth and twenti-
eth centuries was the language of prayer. The Vatican decision to use the vernacu-
lar has opened the way to disagreement over which vernacular. At the end of 2006,
a disagreement between Tamil and Kannada supporters in a parish in Jakkalli, an
ancient city in Karnataka state that is about 2,200 kilometres south of New Delhi,
led to court actions and resulted in the police suspending religious services in the
parish.1
Study of the interaction of language and religion is comparatively recent
(B. Spolsky, 2003b), and the study of religious language policy even more so. This
chapter will therefore be exploratory, focusing on the language practices and
beliefs that develop within the religious domain; it will ask how religious institu-
tions and leaders attempt to exercise authority to modify the language practices
and beliefs about the language of their followers and of others. As with other kinds
of language management, the processes involved may be stated explicitly as rules
about language choice and use, or may be implicit in practice: the practice of
conducting services in only one language, e.g., sets a firm policy line.
In much of the world today, religion remains an important social force. One has
just to open a daily newspaper to read about struggles with fundamentalist Islam in
Egypt or Turkey or Pakistan, or controversy over decisions of the Catholic Church,
66 Bernard Spolsky
or debate over the new US President’s prayer breakfast, to realize this. For adher-
ents, a religious institution is commonly the first social structure outside the family
that aims to influence language use. Western Europe may be ending a period of
secularization, as it comes to grips with the fundamentalism of some of its new
immigrants. Religion is no longer banned in the former Soviet Union. Most Arab
countries are constitutionally Islamic. Nation-states which once separated church
and state are again struggling with religious movements or efforts to assert the
authority of religion in matters of morality and of ethical choice. For many immi-
grants, church or mosque or synagogue remains the principal institution helping
to preserve their heritage language.
In this situation, it seems reasonable to ask how religious institutions and lead-
ers impinge on language practices and beliefs. I first discuss the language policy,
beliefs and management efforts associated with specific religions, and then derive
some general principles. There are two major questions: what success might reli-
gious language managers expect to have in modifying the language practices and
beliefs of their congregants and what support can family language managers expect
from religion and religious organizations? I do not consider in detail those areas
where religion impinges on politics or where governments expect to control the
language of the religious institutions.
As with the family domain, we must be cautious about over generalization.
Accepting Fishman’s (1972) warning about the need to define domains empirically
within each community, we are obviously blurring a great number of distinctions
when we speak about Judaism or Christianity or Islam or any other religion, rather
than concentrating on an individual synagogue or church or mosque. At the same
time, it is tempting to try to look for the larger generalizations.
4.2 Jewish Language Policy
I start with language policy and management in Judaism. With all the changes
in Jewish sociolinguistics over three millennia, the use of Hebrew as language of
sacred text and for prayer has remained consistent, in spite of occasional changes.
While the details continue to be in dispute,2 Jewish language use can be summed
up quickly like this. Up until the Babylonian exile (seventh century BCE), the
common language in Judah was Hebrew. During the exile, in Babylonia and in
occupied Judah, language contact led to plurilingualism. Shortly after the return,
it became customary to accompany the public reading of Hebrew sacred texts by
an Aramaic translation. Aramaic became not just the routine language for legal
contracts, but also the vernacular of those living in close association with gentiles.
Greek was added, the language of settlers who established cities in Palestine
and later of the Greek and Roman governments and their puppets. By the time of
Jesus, Palestine was triglossic, with three languages spoken dominantly in different
parts of the country and a functional division, for Jews, between Aramaic as a
vernacular, Greek for relations with government and Hebrew for religious life
(B. Spolsky, 1983).3
Religious Language Management 67
After the expulsion, Jewish communities in their various exiles continued this
trilingual pattern, developing a Jewish variety (Rabin, 1981) based on a gentile
language for internal community use (Judeo-Greek, Judeo-Aramaic, Judeo-French,
Yiddish, Ladino, Judeo-Venetian, for example), learnt the local vernacular for
dealing with non-Jews, and maintained Hebrew and Talmudic Aramaic as a lan-
guage for prayer and study and for literacy. This pattern, with variations and
changes of community and co-territorial languages as Jews were driven or chose to
emigrate from one country to another, continued more or less until the Enlighten-
ment in the eighteenth century in western Europe. The removal of some external
barriers to civil freedom was accompanied by language changes too. In Germany,
the opening of ghetto gates led many Jews to replace Yiddish with standard German.
There were proposals to substitute German for religious uses of Hebrew. This ten-
dency continued after Jewish immigration to the United States, where many Jews
switched to English for all three functions. At the same time, the successful revival
of Hebrew in Israel led to the loss of earlier Jewish plurilingualism as it replaced
immigrant languages, including most traditional Jewish languages (B. Spolsky and
Shohamy, 1999).
What language management was undertaken by Jewish religious institutions?
Jewish normative law (Halakhah) covers not just relations between man and God,
but also civil and criminal matters (Glinert, 1991). At the same time, with the loss
of Jewish legal autonomy after the destruction of the second Temple, it became ‘a
private matter between observant Jews and their conscience’. Throughout the
ages, there have been continuing debates among observant Jews, so that Jewish law
has been constantly developing. At any point in time, there are differences in detail
within a broad canvas of consensus. There is no single central authority; while
some rabbis may be more respected than others, each can only expect to bind his
followers and observant Jews nowadays remain a minority. Glinert (1991) traces
the varied opinions and changing rulings that deal with language. The Talmud
preferred Hebrew for prayer, but allowed exceptions – during the period of Greek
rule, praying in Greek was permitted under certain circumstances. It agreed that
certain prayers and certain documents (marriage and divorce contracts) should
use Aramaic. It was ambivalent about learning Greek, sometimes banning it as a
language of informers and at other times considering it an ornament for girls.
Glinert (1991) notes first that in the major first compilation of Jewish law edited
in the second century of the Common Era details are given of what must be recited
in ‘The Holy Tongue’ and what may be said in any language. But there is no uni-
formity over time or at any one time. The common pattern has been for Hebrew
to be the normative choice for Jewish ritual. The regular weekly readings from the
Five Books of Moses are in Hebrew, accompanied in some communities by oral
translation into Aramaic.4 Most of the prayers in public and private worship are
also in Hebrew, except for the Kaddish, which has come to play a special role as
the prayer for mourners (Wieselter, 1998), which is in Aramaic. In the Diaspora,
the prayer for the head of state was in the co-territorial language. A sermon was
in the vernacular. Some Jewish communities also switched to the local vernacular
68 Bernard Spolsky
for prayer. This happened in Alexandria at the time of Philo. It was the approach
taken by the Reform movement in Germany and later in the United States.
In spite of the strong preference for Hebrew for ritual and public worship, there
has generally been a willingness to accept the translation of sacred texts into the
vernacular, and it has generally been assumed that the teaching of these texts, at
whatever level, will be in the vernacular. Judaism recognizes the value of transla-
tion in giving believers greater access to the sacred texts, provided that the original
text is preserved and recognized (together with its traditional interpretations) as
the authority.
The status granted to Hebrew had one critical language management outcome:
the need to make sure that children develop proficiency in Hebrew as well as in
their home language. The Talmud laid down the requirement that as soon as a boy
reached the age of five, his father should start to teach him Hebrew.5 In practice,
fathers joined together to set up schools in which their sons6 could be instructed,
and later, Jews living in a town could require other Jews to support the school.
The system seems to have been quite successful, leading to very high standards of
Jewish literacy in the Middle Ages.
The revival of Hebrew by the Zionists at the end of the nineteenth century
raised a number of language policy questions for observant Jewish communities.
Israeli Hebrew, like most revived languages, developed its own pronunciation
markedly different from the many different regional pronunciations of ritual
Hebrew. Before the establishment of the State of Israel, most western Jewish com-
munities continued to use a traditional Ashkenazi pronunciation, but slowly many
accepted the modified Sephardic pronunciation that had become the norm in
Israel. Ultra-orthodox communities resisted this, continuing to use the various
Yiddishized pronunciations that they brought with them from different parts of
East Europe.
One of the principal concerns of the Haredi (ultra-orthodox) Jewish commu-
nities, and in particular of the Hasidic sects among them, has been to maintain
separation from the outside community. This is marked by living in closed neigh-
bourhoods, wearing distinctive clothing, strict observance of dietary laws and also
language practices. Fishman (1966) noted that the most successful two groups in
the United States in preserving their heritage languages were the Amish and the
Hasidim, each of whom attempted to avoid other aspects of the behaviour and
practices of their neighbours. Hasidic groups in the United States, the United
Kingdom and Belgium remain the most committed to the maintenance of Yiddish
in the home. In Israel, while maintaining other aspects of communal separation,
Haredi Jews have in the main shifted to home use of Israeli Hebrew in place of
Yiddish (Baumel, 2002, 2003). Some Hasidic sects, under the influence of their
religious leaders, have been making a major effort to reverse this trend (Isaacs,
1998, 1999).
Requiring ritual reading of Jewish sacred texts in Hebrew led to a special kind
of sacred literacy (B. Spolsky, 1991a). The synagogue Bible reading must be from
a text written by hand on a parchment or vellum scroll. As the texts are written
without punctuation or vowels, a synagogue reader must have learnt the correct
punctuation, vocalization and cantillation of the text and must also know when to
replace a written word with another word laid down in the tradition. Thus, learn-
ing to read requires the moderation of a teacher.
Judaism developed high values for literacy and was successful in maintaining
knowledge of Hebrew during the centuries when it was no longer spoken. The
policy, Wisse (2007: 41) points out, of preserving ‘the Hebrew text as the incor-
ruptible source of their teachings while translating it in their languages of daily
use’ was midway between the Christians for whom the language of the Bible was
unimportant and the Muslims for whom Koran could only be in Arabic; it helped
maintain national cohesion while interacting with the cultural environment.
This summary sketch of Jewish religious language policy provides the basis for
understanding the policy of individual Jewish religious institutions. In the smaller
Jewish communities, where there is only one synagogue, some kind of compromise
has emerged between the wide choice of customary practices and beliefs that had
developed over two millennia of exile. In larger communities, there is a tendency
to establish several synagogues, temples (as Reform Jews call them) or shtibels (the
tiny places of worship preferred by Hasidim), each of which varies according to
the demographic make-up and origin of its congregants. Large establishment
synagogues commonly use the standard co-territorial vernacular for sermons
and announcements; Reform temples use the standard language for many prayers;
shtibels regularly use Yiddish as the accompanying language; and linguistically
marked ethnic synagogues may use a heritage language such as North African
Arabic or French, Yemenite, or English for those parts of the services that are not
conducted in Hebrew.
Jewish religious language management then establishes language practices and
propagates language beliefs that vary from those of the home and modify the lan-
guage practices and beliefs of congregants and especially of their children.
4.3 Language Management in Christianity
Founded by plurilinguals living in a multilingual society, and with a long tradition

of active proselyting throughout the world, Christianity, while also focused on its
sacred texts, has commonly been willing to translate these texts into other lan-
guages. Peters (2003: Vol. II chapter 1) argues that the original Christian texts
were either a collection of sayings or a narrative biography. The biographical texts
became the Gospels of the early churches, so that there was little problem in
moving from the probably original Aramaic to the Greek of the New Testament.
It does not seem to have been important to record the sayings of Jesus in the origi-
nal language or languages.7 The early Christians then were not concerned with a
sacred language, and besides the Greek version, vernacular translations in Egyptian
Coptic, Syrian Aramaic, Latin and Slavic soon appeared.
Christianity became the official religion of the Roman Empire by the end of the
fourth century. The Vulgate translation into Latin of the Bible in the fifth century by
Jerome provided a text which gained almost sacred status during the Middle Ages.
70 Bernard Spolsky
In the eleventh century, Christianity was organizationally divided into the Western
Church, led by the Pope in Rome, and the Eastern Church centred until 1453 in
Constantinople. Latin became the language of the Western Church for all ritual
purposes, maintaining this position until the Second Vatican Council meeting
from 1962 to 1965 permitted the use of the vernacular in the liturgy.
The Eastern Orthodox Church was linguistically pluralistic: with Syriac and
Armenian traditions alongside Greek, it encouraged a Gothic translation in the
fourth century and the Slavonic translation in the ninth (John F. A. Sawyer, 2001).
A typical case was the work of St Stefan of Perm, the fourteenth century Russian
Orthodox bishop responsible for converting the Komi people and developing an
alphabet for them (Ferguson, 1968).
The Western church in contrast maintained a strict language policy. The Span-
ish and Portuguese conquests of South America were partly conceived and oper-
ated as religious activities. The goal of the conversion was the degradation and
destruction of the autochthonous religions and languages. For the Catholic Church,
Latin was the language of ritual and the Latin translation of the Bible was the only
approved version. The catechism could however be taught in the vernacular.
In the sixteenth century, Christianity in western Europe underwent a major
religious and linguistic change with the Protestant Reformation, providing direct
access to the Bible by translation into the vernacular. This was not just a matter of
providing new texts, but a radical case of language management. The Reformation
also involved active iconoclasm, moving from images to words. English churches
during the reign of Henry VIII had their icons destroyed and were required to
purchase English Bibles for their still largely illiterate congregations (E. Spolsky,
2007).
As the Protestant movement fragmented into small proselytizing sects, its mis-
sionaries started to spread ahead of or together with the soldiers and sailors estab-
lishing colonial empires for Western nations. As Sugirtharajah (Sugirtharajah,
2005) put it, the ‘Bible, beer, a gun and a printing press’ were conjoined colonial
artefacts. This process spread religion, colonial rule and literacy, producing major
changes in the sociolinguistic ecology of many parts of the world. As a general rule,
Roman Catholic missionaries – no less colonialist than others, as witnessed in the
conquest of Latin America – were satisfied to learn the languages of their converts
well enough to teach the catechism in it.8 Uncommitted to the sacred status of any
language, Protestant missionaries on the other hand set out to translate the Bible
into the local language.
Protestant missionaries thus played a key role both in the development of liter-
acy in local vernaculars and then at a later stage in the spread of the colonial and
metropolitan languages with which they were associated. A good example is mis-
sionary work among the Polynesians in the South Pacific. In Samoa, New Zealand
and Tonga, English missionaries arriving at the beginning of the nineteenth cen-
tury were successful both in converting the local people to their own version of
Christianity and in establishing strong vernacular literacy. Literacy was introduced
into Tonga in 1829by the Wesleyan missionaries, who set up schools to teach chil-
dren and adults to read and write in the Tongan language (Latukefu, 1974, 1980: 55).
In six months, many chiefs could make themselves understood in English and,
when asked, would write down the names of the islands on a slate. The London
Missionary Society had as its fundamental principle, ‘our design is not to send
Presbyterianism, Independency, Episcopacy, or any other form of church Order
and Government, about which there may be differences of opinion among serious
persons, but the Glorious Gospel of the Blessed God to the Heathen. . . .’ (Garrett,
1982: 10). The translation of the Bible into Tongan therefore was a first major
task. Moreover, when they found many Tongans did not easily learn English, they
accepted the challenge of themselves learning Tongan in order to teach in it. One
of the most prolific of the translators was J. E. Moulton, who translated two vol-
umes of world history, Milton’s Paradise Lost, two volumes of Pilgrim’s Progress, and
a geography of the Holy Land into Tongan. In missionaries and educators such as
Moulton, the Tongans had people willing to let them gain European knowledge in
their own language (B. Spolskyet al., 1983).
Similar approaches in New Zealand led to the rapid development of Māori
vernacular literacy, so that by 1860, there was probably higher literacy among the
indigenous Māori than among the English settlers who had started to arrive in
1840. Missionaries accepted the local language, adding literacy and working
towards language cultivation. Under colonial education policy starting in 1867,
the process was reversed, and serious Māori language loss started (B. Spolsky,
2003a, 2005).
In New Zealand, where the indigenous language appears to have been reasona-
bly homogeneous,9 one effect of Bible translation was to standardize Māori. In
other cases, there was a more radical effect. In Fiji, e.g., the missionaries translated
the Bible into the Bau dialect which as a result became the standard language now
recognized by the school system. For speakers of other dialects, the standard Fijian
of school is as distant from their home language as the standard Hindi of school is
from the many dialects actually spoken by Fijian Indians.
In the Belgian Congo, during the period of King Leopold II’s independent state
(1884 to 1904), the missions were under colonial control, their land granted by the
colonial government. In return, the missions were expected to run the system of
education. The colonial charter issued in 1908 required that French and Flemish
be used for official documents but allowed the option of teaching African lan-
guages in schools. For the next 30 years, the missionaries were active in developing
a large number of the dictionaries and grammars subsidized in part by the colonial
government. There was emphasis on literacy and on the development of mass-
publication, some of it religious in nature but not necessarily so (Fabian, 1983).
As in many other colonies, the missionaries in the Congo became caught in the
struggle between indigenists who aimed to preserve African culture and assimila-
tionists who argued for Europeanization. For missionaries, the principal task of
education was to teach Christianity in the mother tongue. The question of the
variety of mother tongue was critical, and they generally adopted a hierarchical
view, believing that a small number of supraregional languages (with French at
the top) would best serve the needs of the colony. At first, four African languages
(Kikongo, Lingala, Tshiluba and Swahili) were chosen by the colonial administration
72 Bernard Spolsky
in consultation with the missionaries. By 1948, Swahili had become the lingua
franca of eastern Congo described and standardized in the grammars prepared by
missionaries. In Zimbabwe, it was missionary activity that led to the virtual creation
of a standard variety (Ranger, 1989). These two examples show the complex but
important role played by missionaries and their language management activities
in the changes in the sociolinguistic ecology and on the development of ethnic
identities.
Among the Navajo, the fact that only about a third of the people became Christian
had important outcomes. Here, too, there was a distinction between the Roman
Catholic missionaries, who produced an early grammar and dictionary, and the
Protestants whose work in Bible translation had a major influence on the develop-
ment of limited vernacular literacy. The influence was strengthened by association
with the colonial administration, in this case, the Bureau of Indian Affairs. The
decision of the Bible translators to adopt the government orthography developed
in the late 1930s by Bureau specialists was a critical factor in its acceptance (Young,
1977). In the 1940s, those who were literate in Navajo could read the newspaper
which was justifying the war effort, or the New Testament being printed by the
American Bible Society. During the brief flowering of Navajo bilingual education
in the 1970s, there seemed to be agreement between those working to maintain
Navajo through using it in the schools and the Protestant Churches. However, in
many cases, opposition to the teaching of Navajo in the schools came from the
churches, which saw a danger that Navajo culture, meaning the traditional Navajo
religion, would be taught (B. Spolsky, 2002).
The Protestant tradition of Bible translation was the basis of a very wide range
of language management, for in most cases, it required the development of writ-
ten and standardized languages out of what had previously been loosely associated
vernacular dialects. The British and Foreign Bible Society was founded in 1804 in
London with the sole purpose of encouraging wide circulation of the Bible ‘with-
out note or comment’. In its first year at least one portion of the Bible had been
translated into 67 languages; by now, there have been translations into over 2,000
languages. Societies were formed also in other countries. The American Bible
Society founded in 1816 published its first translation (in Delaware) two years later.
It supported work in India and China and the Levant. Founded in 1934 as a sum-
mer training program, SIL International (formerly known as the Summer Institute
of Linguistics) has been involved in training linguists to engage in research and
language development work. Researchers carry out fieldwork studying the lan-
guage and culture of indigenous groups and helping translate the Bible and other
material into the local language. Over the years, SIL has completed work in 400
languages and is active in another 1000. Over 450 translations of the New Testament
and other scripture portions have been published. In the changing atmosphere,
SIL researchers cooperate with national governments and with the speakers of
endangered languages in developing literacy programs. SIL International defines
itself as a ‘non-profit, scientific educational organization of Christian volunteers’.
Religious language policy plays an important role in providing support for the
maintenance of a heritage language. Judaism, by preserving Hebrew as the language
of ritual and adding an educational system to guarantee intergenerational conti-

nuity, effectively kept the language alive for nearly two millennia and provided the
basis for its revival. Christian churches preserve the older form of languages (Old
Church Slavonic, Gothic and Latin). Vernacular church services and associated
educational and social programs play a significant role in supporting family lan-
guage policy in the encouragement of immigrant language maintenance. A signifi-
cant element in what may be labelled Diaspora churches (Polynesian churches in
the United States, Australia and New Zealand; immigrant churches in the United
States and Australia, for example) is that they tend to continue to conduct services
in the immigrant vernacular. B. Spolsky (1991b; B. Spolsky et al., 1983) describes
this for Tongan and Samoan churches. Woods (2002) describes ethnic churches in
Australia. Azarya (1984) describes the maintenance of Armenian in Jerusalem.
Studying migrants in Greece, Gogonas (2007) found that religion played a key
role in the greater language maintenance among Muslim Egyptians than among
Albanian immigrants who lacked a religious tradition.
As may be expected, religious language management often intersects with
political issues. The use of the national language in Protestant churches after the
Reformation was an expression of political policy. In the nineteenth century, in
what were then called the Northwest Provinces (nowadays Lithuania and Belarus),
the Russian government felt itself threatened by Polonization expressed in the
spread of Roman Catholicism in competition with the Orthodox Church, and the
use of Polish in the sermon and other non-Latin parts of the ritual, and attempted
to introduce Russian into the Roman Catholic services to combat Polish influence
(Weeks, 2002). In Friesland, the use of Frisian in church remains controversial, but
a few nationalist ministers have persuaded their local consistory to allow Frisian for
hymns and sermons (Zondag, 1987).
For Quakerism, also known as the Society of Friends, established in England in
the middle of the seventeenth century, rules for language use were important.
Quakers were expected to limit their speech, in worship as well as in normal life.
Their worship meetings were marked by long silences, during which congregants
weighed carefully anything they might say.10 One grammatical feature was the use
of the already archaic second person singular pronoun (thou and thee instead of
you).11 Titles and honorifics were eschewed, clothing was plain, greetings were
avoided (Graves, 2001).
Religious language management commonly concerns itself with the control and
avoidance of certain kinds of speech. Judaism forbade blasphemy, speaking evil of
God. Christianity continued this and included speaking evil of sacred persons or
objects. When Christianity became an official religion, blasphemy became a criminal
offence through the code of Justinian I in the sixth century. After the Reformation
in England, it became part of common law and was applied rigorously until the
1920s (Pickering, 2001).12
Both Christianity and Judaism institutionally have taken many steps to manage
the language of their congregants. This applies not just to the choice of language,
but also to the form of expression. Halakhah (Jewish religious law) includes a set
of rules for euphemistic speech and another for avoiding slander. Halakhah is
74 Bernard Spolsky
determined by a complex set of analyses of biblical and rabbinical statements

and the acceptance by a community or an individual of one specific ruling. In
Christianity, it depends on the organization of the branch of the church, as to
whether an individual or defined body has authority. In essence, these rules are
applied to public speech events associated with religious worship within individual
churches, but they also have serious implications for and effects on the language
of homes and individuals. They are associated with wide levels of society, such as in
the common alliances between missionaries and colonial governments, or when
religion and ethnicity, or religion and nationalism are blended.
4.4 Islamic Language Management
Islam was founded on principles and practices adapted from both Judaism and
Christianity, and like them, its major development was in the Mediterranean area,
so that it can be considered a Western religion like them, certainly in the premod-
ern period.13 In language policy, Islam broke from both its models in its continu-
ing insistence on the higher status of the Classical Arabic in which the Qur’an was
composed. The Qur’an, believed to be the actual word of God, may only be read
or cited in Arabic (Mattock, 2001). Peters (2003) summarizes: ‘the consequence,
then, is that the Qur’an contains the precise words of God, without human inter-
vention or conditioning of any sort, that God had spoken, and Muhammad had
heard and reported, Arabic speech.’ Some Islamic authorities do not permit
translation.
Keeping the sacred texts pure was critical. M. Y. I. H Suleiman (2001) explains
that the development of the Arabic linguistic tradition, starting at the end of the
eighth century, was intended to deal with the phenomenon called lahn (solecism),
the faulty speech of the new converts to Islam during the rapid spread of the reli-
gion and of Arabic, noted not just in ordinary speech but also in recitation of the
Qur’an, where it represented ‘a dangerous interference in the effort to ascertain
of the message of the Qur’an in its capacity as the revealed word of God verbatim’.
Arabic grammar was a pedagogical tool for teaching the language to non-Arab
converts.
The spread of Arabic from the Arabian Peninsula through the Middle East,
North Africa and Spain, was the result of a combination of religious and military
conquest in the sixth and seventh centuries, its conquest of Europe blocked finally
at the battle of Poitiers in France in 733. The Umayyad Empire in the mid-eighth
century already included Spain in the West and was starting to touch India in the
East. By the fourteenth century, the borders had spread south along the African
coast and through the Sahara and East into India; Turkey too was included. By
1500, Spain had been reconquered by the Christians, but Central Asia in the north
and Malaya in the East had become part of the Muslim world. In the fifteenth
century, the Ottoman Empire, having conquered Constantinople, included the
Balkans, Crimea, Turkey and Syria; later, it added Algeria and Egypt.
Linguistically, the spread of Arabic as a lingua franca was uneven; in the Middle
East, it replaced Aramaic in this role. The further regions – Persia, India, Turkey,
the Berber areas of North Africa, the Sudan – continued to use other vernaculars,
but Classical Arabic was the language of religion and the high culture that devel-
oped in the period of the Golden Age of the Abbasid caliphs. In a few hundred
years, however, Arabic replaced its predecessors – the South Arabian languages
and Aramaic and their dialects, and even Coptic in Egypt. It did not however
replace Turkish (an Altaic language) and while it dominated Persian (an Indo-
European language) for a while, a revitalized version of Persian re-emerged in the
tenth century.
Islam continued to spread, south into Africa and east into South East Asia, but
it was often only the religious language and script that were adopted. Thus, we
need to distinguish between countries where Arabic is official – Algeria, Egypt,
Iraq, Jordan, Kuwait, Lebanon, Libya, Mauritania, Morocco, Oman, Saudi Arabia,
Sudan, Syria, Tunisia, United Arab Emirates, Western Sahara and Yemen,14 each
with its distinct local varieties alongside the official Classical standard – and those
where most Muslims speak a different language and have limited knowledge of
Arabic for religious functions.
Mattock (2001) argues that Islam and Classical Arabic form a symbiosis, Arabic
serving as the language of power for Muslims in the Middle East. For Muslims
anywhere, Arabic remains a sacred language. Islam came to South-East Asia peace-
fully, spreading particularly in the island regions like Indonesia while the main-
land generally remained Buddhist. It seems to have been introduced by traders
(Kratz, 2001). Malay, with borrowing from Arabic and Persian, became the main
language of Islamicization.
The 250 million Muslims in South Asia have accepted the sacred primacy of
Arabic since the first communities were established (Shackle, 2001). In many areas,
Arabic persisted for ‘ritual inculcation of a more or less mechanical recognition of
Arabic sufficient for recitation of the Qur’an’ (Shackle, 2001: 63).
While children in Arabic speaking countries have to develop literacy in only
one language, Muslim children in Pakistan and India must learn Arabic for reli-
gious purposes and another language for secular (Rahman, 2006). Traditionally,
children in Pakistan learnt to read in Persian and in Arabic, the Arabic being
taught in the madrasas. Under British rule, Arabic was restricted to religious use
and pupils learnt read the Qur’an but did not learn Arabic. Arabic is nowadays
compulsory in schools in Pakistan; the medium of instruction is generally Urdu.
While some Muslims oppose the teaching of English, most who can afford it want
their children to learn it.
J. F. A. Sawyer (2006) discusses the effect of religion on literacy. He notes cases
where religion has actively obstructed popular access to reading and writing. The
first is the way that the Christian religious establishment in mediaeval Europe tried
to prevent ordinary people from reading the Bible in the vernacular. In India too,
the scripts in which Sanskrit and Hindi are written were considered divine and
only to be used by trained personnel. In Persia, Zoroastrian priests taught that the
sacred text (the Avesta) should not be written down: this, Sawyer said, was reported
also to be characteristic of the Druids. He argued that the spread of Islam also had
a negative effect on literacy. Muslim law forbade translating the Qur’an into the
vernacular (the translations into mediaeval Persian and Ottoman Turkish were
76 Bernard Spolsky
exceptions) and required the teaching of Arabic. Schoolchildren in Qur’an schools

learn to read and write Arabic but not their mother tongue. Arabic script was then
used to write the various vernaculars. The effect of this policy, he argues, was to
slow down the development of literacy in those areas where Islam and the Arabic
language were strongest. Illiteracy, according to current UNESCO statistics, con-
tinues to be higher in the Arab states, sub-Saharan Africa, and South and West
Asia, than in East Asia or Latin America.
Maamouri (1998) considers Arabic diglossia in the Middle East and North
Africa, characterized by growing inadequacy, questionable relevance and unac-
ceptably low level of output. Arabic is important to these countries for identity.
Literacy is acquired late. Conversion meant accepting ‘an elementary form of
Arabic literacy which allowed its users to achieve little more than going through
the daily requirements of the creed’. Over time, a gap developed between the
standardized Arabic of the Qur’an and the ‘corrupt’ spoken language. The higher
status afforded to the written language led to diglossia (Ferguson, 1959; Hudson,
2002). Qur’anic literacy benefited a class of religious professionals but did not
provide functional literacy for the people, Maamouri concludes.
The linguistic effects of Islam have been strong but complex. Where religious
conversion and military conquest were combined, and in areas where the previous
language was Aramaic, a new vernacular emerged, a local spoken variety of Arabic,
but one that was held in contempt by the religious establishment with its funda-
mental commitment to the language of the Qur’an. The battle over possible
acceptance and standardization of the national vernaculars was fought in Egypt in
the 1920, and a combination of Pan-Arabism and Islam won out (M. Y. I. H. Suleiman,
2001; Y. Suleiman, 1996). There are those who argue that these language policy
decisions help account for the failure of Islamic and Arabic countries to adjust to
modernization. In non-Arabic speaking areas, Arabic remained devoted to reli-
gious matters.
4.5 Other Religious Language Management
Hinduism from the religious point of view is better considered as a ‘variety of reli-
gious traditions linked by Indian cultural history and to some extent by the use of
the Sanskrit language’ (Killingley, 2001: 52). Its seeming similarity to the three
major Western religions is largely an artefact of its reinterpretation by Western
Orientalists, scholars and missionaries (King, 1999). The notion of Hinduism as a
religion emerged in the nineteenth century (the 1955 Hindu Marriage Act defines
Hindu as anyone who is not Muslim, Christian, Parsee or Jew, thus including
Buddhists, Jains and Sikhs). One of the results of this Western view was to see
Hinduism as a normative religion; another was to ‘textualize’ it by concentrating
attention on the Sanskrit texts of the Veda. The oral ‘popular’ aspects of Indian
religious tradition were ignored, and decried as not following the texts. This had
major effects on the development of Indian literacy, providing ‘scribal communi-
ties and authoritative interpreters’ that were essential for efficient administration
of colonial India. It was essentially the elitist Brahman forms and traditions that
were stressed. The multiplicity of Indian religious tradition was thus standardized
and unified, in a form consistent with Western models. This development paral-
leled and was related to the growth of Hindu nationalism, which chose to stress the
Sanskrit tradition.
Hindu remained an ethnic term at least until the nineteenth century, when it
picked up its religious and nationalism meanings. In India, language choice often
correlates with a religion: there are regions where Hindus speak Marathi, Muslims
speak Urdu and Jains Kannada. Hindu nationalism promoted Hindi as the modern
language for all India and Sanskrit as the language of scholarship. Opposition to
Hinduism, such as in Tamil Nadu, sometimes took the form of atheism (Killingley,
2001: 53).
The oldest South Asian religious texts, forming the Veda, are in Sanskrit. Sanskrit
is the set language of ritual, and the texts, unwritten until the third century BCE,
were transmitted orally from teacher to pupil. Because of the sacredness of the
mantras, accurate transmission is required. Vedic priests were trained in phonet-
ics, grammar, etymologies and metre. Sanskrit spread from ritual to scholarly use
(Killingley, 2001), and Sanskrit words were borrowed freely into Hindi. But the
texts were regularly translated into other vernaculars, so that other languages
(Kannada, Marathi and English) were widely used. Outside India, Sanskrit is used
in ritual (especially weddings and other life-cycle events), but English glosses
and explanations are common. Contemporary Hindus often use religious texts in
other modern Indian languages. Some Vedic hymns are used in major life cycle
rituals, though participants do not understand them: women and lower caste
Hindus were forbidden to study the Vedas (Brockington, 2001). In South Asia,
Pandharipande (2006) argues, there is no longer a strict equation of religion and
language. In the Middle Ages, content rather than form was accepted as the differ-
entiating feature, and all languages were considered equally able to express the
content. While the Classical languages (Sanskrit, Arabic and Pali) remain con-
nected with a single religion, other South Asian languages, including English, have
begun to be used for the various religions.
4.6 Conclusion
Religious institutions set up practices for those people who are observant; they
ascribe values to varieties of language, and work to establish language policies. An
observant Jew is encouraged to learn Hebrew, and an observant Muslim to acquire
some knowledge of classical Arabic. They can provide support for the maintenance
of heritage languages. With their own internal policy structures – rabbis, priests
and imams passing on beliefs and practices to congregants – they become an
important external factor adding to the pressures on adherents and on their
homes. They establish the religious ideology or ‘ethic of divinity’ (Shwederet al.,
1997) of their adherents, for whom the choice of language is a matter of sacred
tradition – helping explain the strong reactions (akin to disgust) of some Catholics
to the vernacularization of the Mass, of some Jews to the Reform use of German or
English in prayer and of Muslims to the potential impurities (lahn) in the Qur’an.
78 Bernard Spolsky
Translating these generalizations to any specific instance, one must take into
account the communication requirements of the situation. The key participants
in the religious domain are the divinity (to whom prayer is addressed and who is
the author of sacred texts), the congregants and any intermediary minister. In the
various religions and in various parts of the religion, there are different attitudes
to maintaining a single language for sacred texts (communication from the divi-
nity) and for prayers (communication from the congregation to the divinity). As a
general rule, the minister is more likely to know and use this special language, and
when the congregants do not know it, the minister is likely to be expected to trans-
late the sacred text into a language that the congregants do know. Also, as a gen-
eral rule, the minister’s communication with the congregants is most likely to be
in their own language. When the United States took over New Mexico, it replaced
the Spanish-speaking Mexican priests with French-speaking priests from Louisiana,
leading to breakdown in the priests-congregation interaction. In a small synagogue
in a German town, the rabbi (himself English-speaking) told me that the German-
speaking heads of the local community had discouraged his efforts to learn Russian,
the language of former Soviet Jews who regularly attended services. Inability to
communicate with congregants in their vernacular is a handicap for ministers.
Congregants are assumed capable of memorizing or ‘reading’ aloud prayers writ-
ten in a language they do not understand: prayer books may then include transla-
tions alongside the text in the sacred language. Speaking to each other, congregants
will generally use their vernacular.
Attempts to vary these obvious patterns will be evidence of the existence of
external forces. For example, in Friesland, those ministers who attempt to conduct
their services in Frisian rather than in the standard Dutch which has become the
norm are clearly expressing strong activist language ideology. Lebanon too is
an example of the strong influence of religion of language knowledge and use
(Joseph, 2006). Since the conquest in the seventh century, it has been essentially
Arabic speaking, but during the Ottoman period, bilingualism started to emerge.
Government officials of whatever religion were bilingual in Arabic and Turkish.
Christians tended to be bilingual in Arabic and French (the language of their
major Western European protector), with Maronite Christians also maintaining
Syriac as their liturgical language. Up to the end of World War I, anyone who knew
French was likely to be Maronite or Roman Catholic, anyone who knew English
to be an educated Orthodox Christian or Muslim. French-Arabic bilingualism con-
tinued to increase, especially among Christians. French declined after 1960, but
in the recent unrest, is once again being claimed especially by Christians. This
example shows the complex relationship between religion and politics in the
values assigned to languages.
What can one predict for the future? The past decade or so has been marked
by a new recognition of the importance of religion in the world, and some see the
growing fundamentalism of various religions as new social force. One obvious
tendency is towards maintaining the boundaries between religions by insisting on
separate traditional sacred languages. This is of course modified by the efforts of
some religions to convert others. At the same time, globalization and the global
spread of religions provide support for the tendency to move towards a global
language; there are signs of Hinduism and even Islam relaxing their efforts to
maintain the traditional language among adherents who speak other languages.
Time will tell, but one can see good reasons for encouraging continuing and
deeper study of the field of religious language management.
Notes
1 UCA News, 18 January 2007.

2 Towards the end of 2004, arguments about the origin of Yiddish and about the
relationships between pre-revival and revived Hebrew burst into public view again.
3 The position of Latin is less certain. Adams (2003: 608) concludes that it was a
‘super-high’ language in the army.
4 This practice of following each Hebrew sentence from the Torah with an Aramaic
sentence from the Targum continues in Yemenite synagogues today, even though
knowledge of Aramaic is now less common.
5 One assumes that this was after Hebrew stopped being spoken in the home.
6 While there is earlier evidence of educated women, formal religious education for
women started only with the Beis Yaakov movement in the early twentieth century.
7 First century Palestine was multilingual and most of its inhabitants plurilingual.
8 In the process, many of them wrote valuable pioneering grammars and dictionaries.
9 While there were minor phonological and grammatical dialect differences, and
more significant lexical differences, the only major dialect difference was the
variety spoken by the tribe resident in the sparsely populated South Island (Harlow,
2007).
10 Collins (2003) analyses the structure of successful discourse of participants in
Quaker meetings.
11 Birch (1995) reports that the use of the second person singular pronoun is now
rare both in Quaker homes and meeting-houses.
12 Islam too takes blasphemy seriously: witness the case of the fatwa issued in 1989
calling for Salman Rushdie’s death, or the 2007 jailing of an English school teacher
in the Sudan for allowing her pupils to name a teddy bear ‘Mohammed’.
13 See Peters (2003).
14 It is also official in the West Bank and Gaza and alongside Hebrew in Israel.
References
Adams, J. N. (2003), Bilingualism and the Latin Language. Cambridge: Cambridge

University Press.
Azarya, V. (1984), The Armenian Quarter of Jerusalem. Berkeley: University of California
Press.
Baumel, S. D. (2002), ‘Language policies of ethnic minorities as influenced by social,
economic, religious and political concentrates: An examination of Israeli Haredim’.
Unpublished Ph.D., Bar-Ilan University, Ramat-Gan, Israel.
80 Bernard Spolsky
Baumel, S. D. (2003), ‘Teaching English in Israeli Haredi schools’, Language Policy, 2,

(2), 47–67.
Birch, B. M. (1995), Quaker plain speech: A policy of linguistic divergence. Interna-
tional Journal of the Sociology of Language, 116, 39–59.
Brockington, J. L. (2001), ‘Hindu sacred texts’ in J. F. A. Sawyer, J. M. Y. Simpson and
R. E. Asher (eds), Concise Encyclopaedia of Language and Religion. Amsterdam:
Elsevier, pp. 126–127.
Collins, P. (2003), Storying self and others: The construction of narrative identity.
Language and Politics, 2, (2), 243–264.
Fabian, J. (1983), Missions and the colonisation of African languages: Developments in
the former Belgian Congo. Canadian Journal of African Studies, 17, (2), 165–187.
Ferguson, C. A. (1959), ‘Diglossia’, Word, 15, 325–340.
Ferguson, C. A. (1968), ‘St. Stefan of Perm and applied linguistics’, in J. A. Fishman,
C. A. Ferguson and J. D. Gupta (eds), Language Problems of Developing Nations.
New York: Wiley, pp. 253–265.
Fishman, J. A. (1972), Domains and the relationship between micro- and macrosocio-
linguistics, in J. J. Gumperz and D. Hymes (eds), Directions in Sociolinguistics.
New York: Holt Rinehart and Winston, pp. 435–453.
Fishman, J. A. (ed.) (1966), Language Loyalty in the United States: The Maintenance and
Perpetuation of non-English Mother Tongues by American Ethnic and Religious Groups. The
Hague: Mouton.
Garrett, J. (1982), To Live among the Stars. Geneva and Suva: World Council of Churches
and University of the South Pacific.
Glinert, L. (1991), ‘Language choice and the Halakhic speech act’, in R. L. Cooper
and B. Spolsky (eds), The Influence of Language on Culture and Thought: Essays
in Honor of Joshua A. Fishman’s Sixty-Fifth Birthday. Berlin: Mouton de Gruyter,
(pp. 157–182).
Gogonas, N. (2007), Ethnolinguistic Vitality and Language Maintenance in Second-Genera-
tion Migrants: A Study of Albanian and Egyptian Pupils in Athens. Unpublished D.Phil
dissertation, University of Sussex.
Graves, M. P. (2001), ‘Quakerism’, in John F. A. Sawyer and J. M. Y. Simpson (eds),
Concise Encyclopedia of Language and Religion. Amsterdam, New York, Oxford,
Shannon, Singapore, Tokyo: Elsevier, pp. 83–84.
Harlow, R. (2007), Māori: A Linguistic Introduction. Cambridge: Cambridge University Press.
Hudson, A. (2002), ‘Outline of a theory of diglossia’. International Journal of the Sociology
of Language, 157, 1–49.
Isaacs, M. (1998), ‘Yiddish in the orthodox communities of Jerusalem’, in Dov-Ber
Kerler (ed.), Politics of Yiddish: Studies in Language, Literature and Society. Lanham,
MD: Altamira Press, Vol. 4, pp. 85–96.
Isaacs, M. (1999), ‘Contentious partners: Yiddish and Hebrew in Haredi Israel’, Inter-
national Journal of the Sociology of Language, 138, 101–121.
Joseph, J. E. (2006), ‘The shifting role of languages in Lebanese Christian and Muslim
identities’, in T. Omoniyi and J. A. Fishman (eds), Explorations in the Sociology of
Language and Religion. Amsterdam and Philadelphia: John Benjamins Publishing
Company, pp. 165–179.
Killingley, D. D. (2001), ‘Hinduism’, in John F. A. Sawyer and J. M. Y. Simpson (eds),

Shannon, Singapore, Tokyo: Elsevier, (pp. 52–54).
King, R. (1999), ‘Orientalism and the modern myth of “Hinduism”’, Numen, 46, (2),
146–185.
Kratz, E. E. (2001), ‘Islam in Southeast Asia’, in J. F. A. Sawyer and J. M. Y. Simpson
(eds), Concise Encyclopedia of Language and Religion. Amsterdam, New York, Oxford,
Latukefu, S. (1974), Church and State in Tonga (1980 ed.). Canberra, Australia: Australian
National University Press.
Maamouri, M. (1998), Language Education and Human Development: Arabic Diglossia and
its Impact on the Quality of Education in the Arab Region. Paper presented at the The
World Bank: The Mediterranean Development Forum, Marrakech.
Mattock, J. N. (2001), ‘Islam in the Near East’, in J. F. A. Sawyer and J. M. Y. Simpson
Pandharipande, R. V. (2006), ‘Ideology, authority and language choice: Language of
religion in South Asia’, in T. Omoniyi and J. A. Fishman (eds), Explorations in the
Sociology of Language and Religion. Amsterdam and Philadelphia: John Benjamins
Publishing Company, (pp. 141–164).
Peters, F. E. (2003), The Monotheists: Jews, Christians and Muslims in Conflict and Competi-
tion. Princeton, NJ: Princeton University Press.
Pickering, W. S. F. (2001), ‘Blasphemy’, in John F. A. Sawyer and J. M. Y. Simpson (eds),
Shannon, Singapore, Tokyo: Elsevier, p. 240.
Rabin, C. (1981), ‘What constitutes a Jewish language?’ International Journal of the Sociol-
ogy of Language, 30, 19–28.
Rahman, T. (2006), ‘Muslim/Islamic education in Pakistan and India’, in K. Brown
(ed.), Encyclopedia of Languages and Linguistics. Cambridge: Elsevier, Vol 8, Entry
4218, p. 409.
Ranger, T. (1989), ‘Missionaries, migrants and the Manyoka: The invention of ethnicity
in Zimbabwe’, in Leroy Vail (ed.), The Creation of Tribalism in Southern Africa. Berkeley
and Los Angeles: University of California Press.
Sawyer, J. F. A. (2006), ‘Literacy and religion’, in K. Brown (ed.), Encyclopedia of Lan-
guages and Linguistics. Cambridge: Elsevier.
Sawyer, J. F. A. (2001), ‘Christianity in Europe’, in J. F. A. Sawyer and J. M. Y. Simpson
Shackle, C. (2001), ‘Islam in South Asia’, in J. F. A. Sawyer and J. M. Y. Simpson (eds),
Shweder, R. A., Much, N. C., Mahapatra, M., and Park, L. (1997), ‘The “big three” of
morality (autonomy, community, and divinity), and the “big three” explanations
of suffering’, in A. Brandt and P. Rozin (ed.), Morality and Health. New York:
Routledge, pp. 119–169.
82 Bernard Spolsky
Spolsky, B. (1983), ‘Triglossia and literacy in Jewish Palestine of the First Century’.
International Journal of the Sociology of Language (42), 95–110.
Spolsky, Bernard. (1991a). Control and democratization of sacred literacy. In Samuel
Rodin (Ed.), Encounters with Judaism: Jewish Studies in a non-Jewish World (pp. 37–53).
Hamilton: Waikato University and Colcom Press.
Spolsky, B. (1991b), ‘The Samoan language in the New Zealand educational context’.
Vox, 5, 31–36.
Spolsky, B. (2002), ‘Prospects for the survival of the Navajo language: A reconsidera-
tion’. Anthropology & Education Quarterly, 33(2), 1–24.
Spolsky, B. (2003a), Reassessing Maori regeneration. Language in Society, 32(4),
553–578.
Spolsky, B. (2003b), ‘Religion as a site of language contact’. Annual Review of Applied
Linguistics, 23, 81–94.
Spolsky, B. (2005), ‘Maori lost and regained’ in A. Bell, R. Harlow and D. Starks (eds),
Languages of New Zealand (pp. 67–85). Wellington: Victoria University Press.
Spolsky, E. (2007), Word vs. Image: Cognitive Hunger in Shakespeare’s England. London:
Palgrave Macmillan.
Spolsky, B., Engelbrecht, G. and Ortiz, L. (1983), ‘Religious, political, and educational
factors in the development of biliteracy in the Kingdom of Tonga’. Journal of Multi-
lingual and Multicultural Development, 4, (6), 459–470.
Spolsky, B. and Shohamy, E. (1999), The Languages of Israel: Policy, Ideology and Practice.
Clevedon: Multilingual Matters.
Sugirtharajah, R. S. (2005), The Bible and Empire: Postcolonial Explorations. Cambridge:
Cambridge University Press.
Suleiman, M. Y. I. H. (2001), ‘Arabic linguistic tradition’, in J. F. A. Sawyer and
J. M. Y. Simpson (eds), Concise Encyclopedia of Language and Religion. Amsterdam:
Elsevier, pp. 326–336.
Suleiman, Y. (1996), ‘Language and identity in Egyptian nationalism’ in Y. Suleiman
(ed.), Language and Identity in the Middle East and North Africa. London: Curzon
Press, pp. 25–38.
Weeks, T. R. (2002), ‘Religion and Russification: Russian language in the Catholic
Churches of the “Northwest Provinces” after 1863’. Kritika: Explorations in Russian
and Eurasian History, 2(1), 87–110.
Wieselter, L. (1998), Kaddish. New York: Alfred A. Knopf.
Wisse, R. (2007), Jews and Power. New York: Schocken/Nextbook.
Woods, A. (2002, June), The Role of Language in Some Ethnic Churches in Melbourne. Paper
presented at the Colloquium on The Sociology of Language and Religion, University
of Surrey Roehampton.
Young, R. W. (1977), ‘Written Navajo: A Brief History’, in J. A. Fishman (ed.), Advances
in the Creation and Revision of Writing Systems. The Hague and Paris: Mouton,
pp. 459–470.
Zondag, K. (1987), ‘“This morning the church presents comedy”: Some aspects of
Frisian in the religious domain’. International Journal of the Sociology of Language, 64,
71–80.
CHAPTER
5
Language and Culture1
N. J. Enfield
5.1 Introduction: ‘Language’ and ‘Culture’
Language and culture are resources for carrying out the business of social relations.
They are systems for interpreting and regulating social action within the complex,
relationship-grounded style of group living that is characteristic of homo sapiens
(Hinde, 1976; Sacks, 1992). These systems are built upon multiple foundations, of
which three matter most: social intelligence of a uniquely human kind; a special,
relationship-grounded brand of social group organization; and a historically
cumulative frame of community-grounded conventions (Enfield and Levinson,
2006). Both language and culture are embedded in this infrastructure. They both
inhibit and promote the kinds of social goals we may have, and the means we have
for achieving them. Together, language and culture yield a framework through
which individuals pursue social goals, enacting, incrementing and otherwise inhab-
iting positions in a vast network of social statuses (Linton, 1936; Kockelman,
2006b).
Attempts to define the words language and culture have been shaped by, or have
shaped, intellectual fashions of the times (Duranti, 1997; Foley, 1997; Kuper, 1999;
Layton, 1997).2 There are two reasons why scholarly traditions may differ in their
view of what language or culture are. These differences may be substantive, due to
contrasting empirical and analytic bases. Or the differences may merely be rooted
in contrasting sociological and ideological positions. A strong stance on the proper
approach to language or culture may often imply, if not explicitly claim, that other
approaches are irrelevant or wrong. For instance, in linguistics, some brands of
‘functionalism’ might imply that highly formalized accounts of grammar, or even
just accounts of grammar which make explicit reference to cognitive representa-
tions, do not represent psychological reality (Hopper, 1987; cf. Goldberg, 2006).
Or a strong ‘formalist’ position might argue that patterns of language usage have
nothing to do with grammar at all (cf. Newmeyer, 2003). Or in anthropology, a
strong ‘practice’ position may imply or claim that culture cannot be captured in
terms of psychological representations, and so on. But language and culture are
both so multi-faceted that a proper account will have to be heterogeneous: both
psychological and practical, both private and public, both formal and functional.
84 N. J. Enfield
5.2 Language and Culture as Cognitive Processes
The semiotic processes in which language and culture are observable necessarily
involve a dynamic interface between the private domain of cognition and the pub-
lic domains of perception and action. Knowledge, thought and moral value are
indeed private but they are not only private. They also necessarily have a public face.
They partake in processes which are caused by, and give rise to, public phenomena
such as speech acts and other kinds of perceptible behaviour. Conversely, signs,
artefacts and social practices are public, but they also have a private side, since they
partake in processes which are caused by, and give rise to, perceptions, memories,
beliefs, intentions and plans (Kockelman, 2006a). The things I know and think
are brought about by what I have seen and heard. And the things I say and do are
caused by things I know and think. Progress towards a full understanding of lan-
guage and culture will mean abandoning the tendency to dichotomize, to assume
that if one way of looking at the problem is certainly correct, then other views are
mistaken. This just isn’t the case.
Both language and culture are cognitive phenomena, which means that they
are grounded in a capacity for flexibly solving means-ends problems, and a
capacity for using mental representations displaced from an immediate context.
This is captured in the definition of cognition offered by Tomasello and Call
(1997: 8):
The prototype of a cognitive adaptation is a behavioral adaptation in which

perceptual and behavioral processes (1) are organized flexibly, with the individ-
ual organism making decisions among possible courses of action based on an
assessment of the current situation in relation to its current goal; and (2) involve
some kind of mental representation that goes ‘beyond the information given’ to
direct perception. Complexity in the decisions to be made is also characteristic
of the prototype of a cognitive adaptation.
Of these two definitive components of cognition, the latter, mental represen-

tation, has been dominant in research on language. The classic Saussurean
sign – signifier standing for signified – embodies this. A vast amount of linguistic
research has aimed at characterizing the content of linguistic representations,
an objective particularly explicit in technical approaches to semantic analysis.
Characteristic of the reflexive nature of language and related types of system, these
approaches explicate the content of linguistic representations with the aid of
linguistic and diagrammatic devices. Paraphrase or definitional approaches may
use different metalanguages and posit different primitives (cf. Wierzbicka, 1980
versus Jackendoff, 1983), but the basic modus operandi is the same: Linguistic
representations are analysed in linguistic or quasi-linguistic terms. More recent
developments in the analysis of meaning in language also invoke non-linguistic
imagery in the explication of linguistic representations (e.g., Lakoff, 1987;
Langacker, 1987).
Language and Culture 85
5.3 Social Cognition and Relationship Thinking
To say that language and culture are cognitive phenomena is not to constrain them
to what is in the head. Like other psychological processes, they are distributed.
They hook into public artefactual material such as symbolic structures (writing,
charts, tables, etc.) and material objects (tools, instruments, environmental struc-
tures; Norman, 1991; C. Goodwin, 1994, 2006; Hutchins, 1995, 2006; Enfield, 2009,
inter alia.). Both language and culture are externally embodied, in symbols, in
behaviour and in material culture. Classical topics in anthropology have this kind
of distributed nature: think of kinship (in ritual events, in reproduction, in daily
social behaviour, in language), biological taxonomy (in the physical world, in liveli-
hood activities, in language), political organization (in social behaviour, in spatial
arrangements, in language).
Beyond the mere fact of our cognitive processes being distributed (in the sense
of Hutchins, 2006), the dimension of human cognition which matters crucially
for the very possibility of culture and language even existing is not shared with any
other species in quite the same form. This has recently been referred to as shared
intentionality (Tomasello et al., 2005), a type of social cognition, grounded in
a special awareness of, and attention to, human relationships, at several levels of
granularity. Accordingly, a proper analysis of both language and culture requires
relationship thinking (Hinde, 1976, 1997; Ingold, 1990). By this I mean both a
kind of thinking that analysts should apply when trying to understand human
interaction, and a kind of cognition that fosters human interaction and its most
prominent machinery, i.e., the words and grammar of language, the practices and
artefacts of culture. I agree with Ingold (1990) on the idea of culture as a ‘logic
of relationships’ (225). This is in line with the views of biologists such as Dunbar
(1988: 2) and Hinde (1976, 1982, 1991), and early comparative anthropologists
like Linton (1936).3
At the core of our social world is the maintenance of relationships entailed by
living in a special kind of social system (Linton, 1936: 113; de Waal and Tyack,
2003; see below). As Linton (1936) outlined, social statuses are ‘polar’ in that they
define relations between people. These relationships are not simply dyadic. What
is special about how humans and other higher primates deal with relationships is
that we are capable of cognitively representing not just the dyadic relationships
that we enter into with others, but the dyadic relationships between others, and
further, in second-order terms, how those relationships between others stand with
reference to our own relationships with those others. Once we recognize the capac-
ity to represent not just relationships but relations between relationships (how one
relationship is related to another relationship, and what that tells us), we derive
powerful modes of thinking about meaning, and therefore, about language and
culture (Kockelman, 2005).
A relationship thinking approach takes communicative interactions as a key
locus (Hinde, 1976; Dunbar, 1988: 12 and passim). Each interaction enacts a spe-
cific, token relationship (e.g. between me and my brother Matt) as well as a type of
86 N. J. Enfield
relationship (e.g. between a man and his brother). Types of relationship in turn
define types of social statuses and identities which will be defining elements
of higher-level social structure (Linton, 1936: 113–131, Radcliffe-Brown, 1952;
Lévi-Strauss, 1953; Nadel, 1957; Hinde, 1976; Sacks, 1992; Dunbar and Spoors,
1995; Hill and Dunbar, 2003; Pomerantz and Mandelbaum, 2005; Enfield and
Levinson, 2006).
5.4 Relationships and Social Structure
Higher-level or second-order social structure emerges out of the interplay of both

negatively and positively valenced forces of relationship-grounded social behaviour.
On the one hand, there are positive pro-social instincts that license trust, compas-
sion and common identity (Henrich et al., 2004; Boyd and Richerson, 2006). On
the other hand are the Machiavellian instincts that license competition, decep-
tion, and social distinction (Byrne and Whiten, 1988; Whiten and Byrne, 1997). In
social interaction, humans are not only interested in (ritually or otherwise) reducing
damage and promoting bonding (Huxley, 1966: 258), but are often equally inter-
ested in marking boundaries and establishing social difference (M. H. Goodwin,
1990: 141–225, 2006; cf. Goffman, 1959, 1967). However, these ostensibly contras-
tive forces are not always easily distinguished: in one frame, an act of altruism
(e.g., spending time and money helping a stranger) can be seen in another frame
as selfish (depriving those closer to you of valuable resources). How it looks depends
on which social unit of analysis we take – the individual? the dyad? the triad? the
family? the ethnic group? Each would be relevant in different contexts.
The idea of relationship thinking is for the analyst to regard human social rela-
tionships as a key locus of analysis of language and culture. While at higher levels
of abstraction, linguistic and social structure can vary enormously across cultures,
the fundamental site at which we observe the development and maintenance of
such structure is in the co-present interactions by which (types of) relationships
are concretely enacted (Hinde, 1976; Dunbar, 1988), and in the special cognition
by which we are able to represent and process these relationships and the relations
between them (Byrne and Whiten, 1988; Kockelman, 2005). The (types of) rela-
tionships enacted in interaction co-define the roles and identities that will ulti-
mately define the sociology and ethnography of a community.
The relevant relationship types may be of two broad varieties, called externally
grounded and reciprocally grounded. (These are not mutually exclusive.) If A and
B stand in an externally grounded relationship, then their relationship is defined
by how they each stand towards some common reference point (with associated
definitive commitments and entitlements). For instance, if A and B are both mem-
bers of a local cricket club, this is a potential basis for defining a relationship
between them through external grounding. Such relations may be negatively
defined as well, where A and B stand differently towards a common reference
point. This type of relationship is also referred to as segmentary (Evans-Pritchard,
1940). Note that in an externally grounded relationship, the relationship between
A and B is not a necessary consequence of their each standing in that way to the
external ground. By contrast, if A and B stand in a reciprocally grounded relation-

ship, then the rights and responsibilities associated with A and B’s incumbency in
that relationship are mutually defined: e.g., if A is B’s father, B is necessarily A’s
child.
Humans are among many species whose behaviour is organized around what de
Waal and Tyack (2003) call individualized, longitudinal society. By describing
human society as ‘individualized’, they mean that ‘members recognize each other
individually and form variable relationships built on histories of interaction’ (de
Waal and Tyack, 2003: x). Importantly, this is independent of any notion of indi-
vidualism as a cultural value or ideology. What is common to all cultures is that
society is made up of distinct, mobile, mortals, who are not telepathic, and whose
interactions must therefore be managed by semiotic means. That is, manipulation
of others in the social world involves the use of signs as tools to cause others’ minds
and bodies to be affected in relatively predictable ways, to relatively predictable
ends. Local ideologies of the relation between person and society are distinct from
this general fact, yet may be constrained by it.
The second property of socially complex societies which De Waal and Tyack
pick out is that they are ‘longitudinal’ (or ‘longitudinally stable’). In a longitudinal
society, ‘species with long life spans have long-term or multigenerational relation-
ships, such as those between grandparents and grand-offspring or friendships
among adults going back to youth’ (de Waal and Tyack, 2003: x; cf. Dunbar,
1988).
I adopt the perspective proposed by de Waal and Tyack, but I use the term
relationship-grounded instead of individualized, to more accurately capture the
idea (cf. Hinde, 1976). Life in a relationship-grounded society presents each indi-
vidual member with a common set of problems of social life. At some level and to
some degree, many of these social problems (and possibly their best solutions)
may be shared with creatures of other relationship-grounded societies such as
those of elephants, bottlenose dolphins, spotted hyenas, baboons and capuchin
monkeys (Dunbar, 1988; de Waal and Tyack, 2003; Sussman and Chapman, 2004).4
Of course, we humans have our own species-unique problems and solutions, but
we are still part of the biological world and this should never be forgotten (Hinde,
1982, 1991; Boesch, 2007, inter alia).
5.5 Social Intelligence
Participants in any interaction are in a culturally and historically specific context,

and in a particular kind of social world, as defined, in part, by species-specific
determinants such as pro-social instincts, social intelligence capacities and struc-
tural constraints on social group size and relationship intensity (cf. Whiten and
Byrne, 1997; Hill and Dunbar, 2003; Richerson and Boyd, 2005). But individual
participants are at the same time mobile agents in distributed populations, each
with their own properties as individuals, each with their own bodies and minds.
Complex social life demands (and enables) complex social cognition (Jolly, 1966;
Humphrey, 1976; Byrne and Whiten, 1988; Tomasello, 1999; de Waal and Tyack,
88 N. J. Enfield
2003; Carpendale and Lewis, 2006, inter alia). People are equipped with a rich
suite of cognitive capacities for navigating the social world, which for convenience
may be referred to as social intelligence. This is not a single capacity or faculty, but
a cluster of related abilities. Consider some of the cognitive capacities that differ-
ent research traditions have focused on, suggesting the kind of social intelligence
that any model of language and culture must presuppose (Carruthers and Smith,
1996; Carpendale and Lewis, 2006; Enfield and Levinson, 2006, inter alia):5
z perspective-taking (awareness of others’ perceptual states)

z false belief understanding (truth vs. people’s representations of it)
z pro-social instincts (altruism, group living, ethnic co-membership)
z cooperative instincts (capacity for flexible joint action toward a mutual goal)
z Machiavellian instincts (dominance, coalition-building, manipulation, ethnic
distinction)
z intention-recognition (attribution of knowledge, belief, desires)
z an intentional stance (intention-attribution to the non-mental realm)
z management and exploitation of mutual knowledge (Schelling thinking)
z a fluid symbolic capacity (sensitivity to social convention)
z docile cultural instincts (propensity to adopt the norms of one’s group)
z socially anchored emotional and moral instincts (motives to adhere to and regu-
late social norms)
These are presumed to reflect universal human capacities, definitive of the cog-
nitive style of our species, and prerequisite for language and culture. But there has
been little serious testing of their robustness across cultures (i.e. whether these
capacities are generically present in individuals in any community), and next to
nothing is known of their cultural permeability (i.e. whether differences in cultural
or linguistic setting may affect the development of such capacities in children).
Linguistic and cultural inflection of social intelligence is a matter for empirical
research.
The importance of social intelligence in language and culture is its role in the
interpretation of others’ communicative actions. Communication is a species of
social action which involves interdependent processes of assessment and manage-
ment (Krebs and Dawkins, 1984; Owings and Morton, 1998). Utterances and their
equivalents are ways of bringing about effects on the world, both in the celebrated
sense of transforming ‘official’ social statuses as in formal rites of passage (Austin,
1962) and in the more workaday processes of transforming mental states, as all
signs do (Kockelman, 2005, 2006a, 2006b). Any individual has capacities to assess
their environment, i.e. to perceptually explore their surroundings and thereby
know new things of consequence (e.g. what to pursue, what to avoid). Individuals
also act upon or manage their environment. One way of managing the environ-
ment is by brute force wielded upon physical objects (say, chopping wood for fire).
In a social setting, however, the most important components of our environment
are other people (cf. C. Goodwin, 2006). When people use language, they are
using controlled signifying behaviour in order to manage their environment by
bringing about effects upon the mental states (intentional states, emotions, habitus,
etc.) of social associates. Like all relationship-grounded creatures, our social action
employs ritualized means of communication (Huxley, 1966) which affect the world
by causing changes in others’ inner states (as opposed to actions which have effects
on the world by brute force; cf. Searle, 1969). In managing the social environment
in this way, a ‘sender’ presupposes and exploits other individuals’ strategies for
management of the social environment (Krebs and Dawkins, 1984; Owings and
Morton, 1998).
As analysts, we therefore want to have a clear sense of what these exploitable
strategies of assessment are (Enfield, in press). They will include social intelligence
capacities along the lines discussed above. These are powerful means for assessing
the social world, tools for ‘reading’ other minds (Byrne and Whiten, 1988; Baron-
Cohen, 1995; Enfield and Levinson 2006, inter alia).6 In the case of humans, the
presupposed capacities for assessment will also include massive second-order
knowledge of the structured semiotic systems known as grammar and culture.
5.6 Social Norms and Interpretive Heuristics
Language and culture are both built on social norms. Norms are learned patterns
of behaviour which are consistent in a community not because it is explicitly stated
anywhere that they be followed like rules, but because not behaving in a manner
consistent with those patterns will be taken as marked, and will attract special
attention in the form of surprise or sanction (Wittgenstein, 1953; Garfinkel, 1967;
Brandom, 1979; Kockelman, 2006b). Social agency is built on this kind of norm-
regulated semiotic commitment, defined as ‘the degree to which one anticipates
an interpretant [i.e., a meaningful response; NJE], where this anticipation is
evinced in being surprised by, and/or disposed to sanction, non-anticipated inter-
pretants’ (Kockelman, 2007: 380). Many patterns of language structure and usage
are like this, including semantics and grammar. More obviously, the standards of
culture are invisible only until they are transgressed (Enfield, 2007).
By contrast, with locally conventional norms, heuristics are logical principles
of interpretation which may be generically applied in attributing meaning to
tokens of communicative behaviour in specific contexts. This is the basic insight
behind Grice’s work on meaning (1957) and conversational inference (1975) (cf.
also Goffman’s idea of framing; 1974, 1981). Grice’s insight can be extracted quite
apart from any ethnocentrism of his widely maligned conversational maxims
(Wierzbicka, 1987, Goddard, 2006). The essential point is as Levinson (2000) puts
it, that amplicative enrichment (Grice’s implicature) is a smart solution to a thorny
bottleneck problem in human communication: we think fast but we speak slow.
While Grice’s (or Sperber and Wilson’s, or Levinson’s) claims about particular
examples may be quibbled with case by case (Wierzbicka, 1991), the principle is
robust: in all cultures, people say more than is said (or convey more than is coded).
That is, interpreters of their communicative actions are able to extract more than
is simply encoded in the conventions of the semiotic resources deployed (e.g. the
dictionary meanings of their words and grammatical constructions). What differs
90 N. J. Enfield
culturally are the local premises (norms) which feed the process of inference, not
the inferential processes themselves (cf. Enfield, 2002, Evans, 2003).
5.7 Language and Culture in Problem-Solving
From this perspective, we can characterize language and culture as tools and con-
texts for communicative action in management of the social world. The domain of
pragmatics, where language and culture come together, may be defined as social
problem-solving. Consider what is meant by problem-solving in a more general
sense.7 If I don’t eat every day I get hungry. To solve this, I might engage in a com-
plex cycle of agricultural practice aimed at harvesting enough rice to meet my
yearly needs, along with hunting and gathering activities to supplement my staple.
Or I need shelter from the weather. To solve this I might build a house.8 Many
imperatives are imposed by genetic and terrestrial fate, and are thus faced in all
cultures. But some imperatives are caused by culture-specific facts. Notice how
problems and solutions are nested. Once I have committed to a certain solution,
this raises new imperatives (cf. Dunbar, 1988: 26–28). For example, as a profes-
sional in a western European socio-economic world, having committed to certain
solutions for the problems of food and shelter, I cannot get along without money.
So, getting money is a solution which in itself becomes an imperative, a problem-
in-need-of-solution. Solutions or strategies will differ widely from human group to
human group.9 Resources for problem-solving include natural materials (e.g.
products of the forest around my village) and culturally acquired tools, instruments
and social conventions.
In solving problems in the social realm, our most important resources are semi-
otic ones – especially, the historically acquired tools which comprise the expressive
resources of any language, along with our social associates and their normative
habits of interpretation. We presume that our social associates will have complex
powers of assessment (outlined above). Our deployment of expressive resources
exploits these powers of assessment as a way of socially managing others in order
to bring about the results we desire. For example, I might combine words into
utterances, and combine these utterances with the transfer of pieces of paper or
coin in order to stave my hunger for the evening. Or I might take my machete to
the forest and return with lengths of wood, bamboo, rattan and palm leaf to repair
my broken house (usually with the help of neighbours). In both cases, I would
typically count my social associates – other people – among my problem-solving
resources.
5.8 Two Imperatives: Information and Affiliation
At least two imperatives can be argued to apply at all times in social interaction,
and are likely to be universal. These are an informational imperative and an affili-
ational imperative (Enfield 2006).10 The informational imperative is to ensure
that our attempts at converging with others on symbolic reference are tolerably
successful (Clark, 1996). That is, we must ensure that we are being understood
by others to a degree sufficient for current communicative purposes. At the same
time, the affiliational imperative is to ensure we are appropriately managing
the social consequences of any interaction (Goffman, 1959, 1967; Heritage and
Raymond, 2005). Every interaction increments an interpersonal relationship by
means of building common experience, and displays the nature of that relation-
ship such that it may be evaluated by participants and onlookers (Enfield, 2006).
A relationship thinking approach puts this in the foreground.
We might also refer to the affiliational imperative as micro-political or coali-
tional, in so far as it has to do with establishing the desired relationships, putting
the other person in or out of some social circle. We are not just generically subject
to an unceasing relationship-consequentiality of social behaviour (and hence
obliged to attend to ritual requirements of face; Goffman, 1959, 1967), we are also
compelled to maintain relationships of certain proximity types (Hill and Dunbar,
2003). The resultant social structure is an outcome of specific cognitive constraints
and a trade-off of numbers of relationships one maintains against time it takes
to service them (Dunbar, 1996). Reality is more textured than this thanks to the
complexities of sociometry (Rogers, 1995; Enfield, 2003, 2005), by which different
individuals solve the trade-off in different ways (distinguishing between, say, ‘weak
ties connectors’ and ‘strong ties homebodies’; Granovetter, 1973, 1978). And not
everyone is equally adept in matters of affiliation and coalition.
5.9 Conclusion
The theme of this chapter has been that language and culture are deeply impli-
cated in a wide set of common human capacities and social functions. Culture can
hardly be learned or enacted without the use of language. And to a great degree,
our linguistic practices define our cultural practices. As many have argued, culture
is widely enacted in talk (Hymes, 1964; Bauman and Sherzer, 1974; Hanks, 1990,
1996; Sacks, 1992; Wierzbicka, 1992; Sidnell, 2005; inter alia). At the same time,
language cannot adequately be described without a framing set of cultural norms
and background. They are learned together, and are vastly co-defining and over-
lapping. While one may ask how language and culture are related, this may wrongly
presuppose that they are separable at all (Hill and Mannheim, 1992; Lee, 1996;
Enfield, 2000). Instead we may ask: What is the common stuff of which language
and culture are built? The answer: They are both cognitive and practical resources
for meaningful action on, and through, social relationships.
Notes
1 This chapter incorporates revised sections of an article titled ‘Relationship think-

ing and human pragmatics’, published by Elsevier in Journal of Pragmatics. I am
92 N. J. Enfield
grateful to the publisher for permission to reprint revised sections of the article.
I am also grateful to Bill Hanks and Paul Kockelman for valuable contributions in
discussion, and to Li Wei for advice and support.
2 Any investigation of language or culture has to cope with their reflexive nature.
Because language and culture are systems of meaning; they can be the objects of
their own meaning-making (Jakobson, 1971; Silverstein, 1976). This unique prop-
erty of human systems of meaning gives rise to an all-pervasive trap in the analysis
of language and culture: the danger of taking local ideology about systems of mean-
ing to be equal to the reality of these systems (cf. Diller and Khanittanan, 2002).
So, analysts of both language and culture need to carefully monitor the distinction
between members’ ideas about their own behaviour, and their behaviour as actually
observed. A linguist must ask: Am I describing language? Or is this language about
language? Am I describing how people talk, or what they say about how they talk?
Similarly for culture.
3 Unlike Ingold I do not see this as incompatible with population thinking (a con-
cept attributed to Darwin: Mayr, 1964: xix–xx; 1970; 1982: 45–47; see Hinde, 1991:
585–586).
4 This does not apply to other complex societies such as those of the ants, since they
are not ‘individualized’ in de Waal and Tyack’s sense.
5 These are not necessarily qualitatively distinct. The list merely represents a range of
different angles on social intelligence from a range of disciplinary traditions.
6 These are also applied in interpretations of the non-social world (Lévi-Strauss,
1966; Goody, 1995; Atran, 2002).
7 I will sometimes distinguish between imperatives as problems which demand
solutions, and strategies as the particular solutions chosen (Dunbar, 1988).
8 See Schutz (1970) and predecessors for a distinction between the ‘because
motives’ which focus on the states of affairs which give rise to actions (I’m picking
berries because I’m hungry) and the ‘in-order-to motives’ which focus on the goals
of actions, or the states of affairs which actions will give rise to (I’m picking berries
in order to eat them).
9 Among the set of problems-in-need-of-solution, some will be generically present
across cultures (e.g. the need to deal with significant problems of speaking, hear-
ing and understanding in conversation; Schegloff, 2006). Others will be present by
virtue of culturally distinct factors (e.g. language-particular thinking-for-speaking
effects, locally specified requirements of politeness, etc.). That is, some features
of code and pragmatics are solutions to problems, some are problems in need of
solution. Culture is always a system for solving problems of social life. It’s just that
some of our problems are caused by the solutions we (habitually) choose and by
the nature of our problem-solving resources – i.e. by culture itself.
10 These are akin to Goffman’s system versus ritual constraints in face-to-face interac-
tion (Goffman, 1981). Paul Kockelman (personal communication) points out that
these correspond roughly to Jakobson’s referential and phatic functions of lan-
guage, two among his six general functions (the other four being emotive, poetic,
conative and metalingual; Jakobson, 1960).
References
Atran, S. (2002), In Gods We Trust: The Evolutionary Landscape of Religion. New York:
Oxford University Press.
Austin, J. L. (1962), How To Do Things with Words. Cambridge, MA: Harvard University
Press.
Baron-Cohen, S. (1995), Mindblindness: An Essay on Autism and Theory of Mind.
Cambridge, MA/London: MIT Press.
Bauman, R. and Sherzer, J. (eds) (1974), Explorations in the Ethnography of Speaking.
Cambridge: Cambridge University Press.
Boesch, C. (2007). ‘What makes us human (Homo sapiens)? The challenge of cogni-
tive cross-species comparison, Journal of Comparative Psychology, 121, (227–240).
Boyd, R. and Richerson, P. J. (2006), ‘Culture and the evolution of the human social
instincts’, in N. J. Enfield and Stephen C. Levinson (eds), Roots of Human Sociality:
Culture, Cognition, and Interaction. London: Berg, pp. 453–477.
Brandom, R. B. (1979), ‘Freedom and constraint by norms’. American Philosophical
Quarterly, 16, (3), 187–196.
Byrne, R. W. and Whiten, A. (eds) (1988), Machiavellian Intelligence: Social Expertise and
the Evolution of Intellect in Monkeys, Apes, and Humans. Oxford: Clarendon Press.
Carpendale, J. and Lewis, C. (2006), How Children Develop Social Understanding. London:
Blackwell.
Carruthers, P. and Smith, P. K. (eds) (1996), Theories of Theories of Mind. Cambridge:
Clark, H. H. (1996), Using Language. Cambridge: Cambridge University Press.
de Waal, F. B. M. and Tyack, P. L. (eds) (2003). Animal Social Complexity: Intelligence, Culture,
and Individualized Societies. Cambridge, MA and London: Harvard University Press.
Diller, A. V. N. and Wilaiwan, K. (2002), ‘Syntactic inquiry as a cultural activity’, in
N. J. Enfield (ed.) Ethnosyntax: Explorations in Grammar and Culture. Oxford: Oxford
University Press, pp. 31–51.
Dunbar, R. I. M. (1988), Primate Social Systems. London and Sydney: Croom Helm.
Dunbar, R. I. M. (1996), Grooming, Gossip and the Evolution of Language. London: Faber
and Faber.
Duranti, A. (1997), Linguistic Anthropology. Cambridge: Cambridge University Press.
Dunbar, R. I. M., and Spoors, M. (1995), ‘Social Networks, Support Cliques, and
Kinship’. Human Nature, 6, 273–290.
Enfield, N. J. (2000), ‘On Linguocentrism’, in Explorations in Linguistic Relativity.
Martin Pütz and Marjolijn H. Verspoor (eds), Amsterdam/Philadelphia: John
Benjamins, pp. 125–57.
Enfield, N. J. (ed.) (2002), Ethnosyntax: Explorations in Grammar and Culture. Oxford:
Oxford University Press.
Enfield, N. J. (2003), Linguistic Epidemiology: Semantics and Grammar of Language Contact
in Mainland Southeast Asia. London: RoutledgeCurzon.
Enfield, N. J. (2005), ‘Areal Linguistics and Mainland Southeast Asia’. Annual Review of
Anthropology, 34, 181–206.
94 N. J. Enfield
Enfield, N. J. (2006), ‘Social consequences of common ground’, in N. J. Enfield and

Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition, and Interaction.
Oxford: Berg, pp. 399–430.
Enfield, N. J., (2007), ‘Meanings of the unmarked: How “default” person reference
does more than just refer’, in N. J. Enfield and Tanya Stivers (eds), Person Reference
in Interaction: Linguistic, Cultural, and Social Perspectives. Cambridge: Cambridge
Enfield, N. J. (2009), The Anatomy of Meaning: Speech, Gesture, and Composite Utterances.
Cambridge University Press: Cambridge.
Enfield, N. J. and Levinson, S. C. (eds) (2006), Roots of Human Sociality: Culture, Cognition,
and Interaction. London: Berg.
Enfield, N. J. (in press), ‘The Elements of Formulation’, in C. Goodwin, C. LeBaron
and J. Streeck (eds), Multimodality and Human Activity: Research on Human Behavior,
Action, and Communication. Cambridge: Cambridge University Press.
Evans, N. D. (2003), ‘Context, Culture, and Structuration in the Languages of Australia’,
Annual Review of Anthropology, 32, 13–40.
Evans-Pritchard, E. E. (1940), The Nuer: A Description of the Modes of Livelihood and Political
Institutions of a Nilotic People. New York: Oxford University Press.
Foley, W. A. (1997), Anthropological Linguistics: An Introduction. London: Blackwell.
Garfinkel, H. (1967), Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Goddard, C. (ed.) (2006), Ethnopragmatics: Understanding Discourse in Cultural Context.
Berlin: Mouton de Gruyter.
Goffman, E. (1959), The Presentation of Self in Everyday Life. New York: Doubleday.
Goffman, E. (1967), Interaction Ritual. Harmondsworth: Penguin Books.
Goffman, E. (1974), Frame Analysis: An Essay on the Organization of Experience. Boston:
Northeastern University Press.
Goffman, E. (1981), Forms of Talk. Philadelphia: University of Pennsylvania Press.
Goldberg, A. E. (2006), Constructions at Work: The Nature of Generalization in Language.
Oxford: Oxford University Press.
Goodwin, C. (1994), ‘Professional Vision’, American Anthropologist 96, (3), 606–633.
Goodwin, C. (2006), ‘Human sociality as mutual orientation in a rich interactive
environment: Multimodal utterances and pointing in aphasia’, in: N. J. Enfield and
Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition, and Interaction.
London: Berg, pp. 97–125.
Goodwin, M. H. (1990), He-Said-She-Said: Talk as Social Organization among Black Children.
Bloomington and Indianapolis: Indiana University Press.
Goodwin, M. H. (2006), The Hidden Life of Girls: Games of Stance, Status, and Exclusion.
Malden: Blackwell.
Goody, E. (ed.) (1995), Social Intelligence and Interaction. Cambridge: Cambridge Univer-
sity Press.
Granovetter, M. (1973), ‘The Strength of Weak Ties’, American Journal of Sociology, 78,
1360–1380.
Granovetter, M. (1978), ‘Threshold models of collective behaviour’, American Journal of
Sociology, 83, 1420–1443.
Grice, H. P. (1957), ‘Meaning’, Philosophical Review, 67, 377–388. (Reprinted in Grice,

1989.)
Grice, H. P. (1975), ‘Logic and conversation’, in P. Cole and J. L. Morgan (eds), Speech
Acts. New York: Academic Press, pp. 41–58. (Reprinted in Grice, 1989.)
Hanks, W. F. (1990), Referential Practice: Language and Lived Space among the Maya.
Chicago: University of Chicago Press.
Hanks, W. F. (1996), Language and Communicative Practices. Boulder: Westview Press.
Henrich, J. et al. (eds) (2004), Foundations of Human Sociality: Economic Experiments and
Ethnographic Evidence from Fifteen Small-Scale Societies. Oxford: Oxford University
Press.
Heritage, J. and Raymond, G. (2005), ‘The terms of agreement: Indexing epistemic
authority and subordination in talk-in-interaction’, Social Psychology Quarterly, 68,
15–38.
Hill, J. H. and Bruce M. (1992), ‘Language and world view’, Annual Review of Anthro-
pology, 21,381–406.
Hill, R. A. and Dunbar, R. I. M. (2003), ‘Social network size in humans’, Human Nature,
14, 53–72.
Hinde, R. A. (1976), ‘Interactions, relationships, and social structure’, Man (New Series),
11, (1), 1–17.
Hinde, R. A. (1982), Ethology: Its Nature and Relations with Other Sciences. London:
Fontana.
Hinde, R. A. (1991), ‘A biologist looks at anthropology’, Journal of the Royal Anthropolog-
ical Institute, 26, (4), 583–608.
Hinde, R. A. (1997), Relationships: A Dialectical Perspective. Hove: Psychology Press.
Hopper, P. J. (1987), ‘Emergent grammar’, in J. Aske, N. Beery, L. Michaelis and
H. Filip (eds) Proceedings of the 13th Annual Meeting of the Berkeley Linguistics Society.
Berkeley: Berkeley Linguistics Society.
Humphrey, N. K. (1976), ‘The social function of intellect’, in P. P. G. Bateson and
Robert A. Hinde (eds), Growing Points in Ethology. Cambridge: Cambridge University
Press, pp. 303–321.
Hutchins, E. (1995), Cognition in the Wild. Cambridge, MA: MIT Press.
Hutchins, E. (2006), ‘The distributed cognition perspective on human interaction’, in
N. J. Enfield and S. C. Levinson (eds), Roots of Human Sociality: Culture, Cognition and
Interaction. Oxford: Berg, pp. 375–398.
Huxley, Sir Julian (ed.) (1966), A Discussion on Ritualization of Behavior in Animals and
Man. Special issue of Philosophical Transactions of the Royal Society of London, Series B,
Biological Sciences, 251(772).
Hymes, D. H. (ed.) (1964), Language in Culture and Society: A Reader in Linguistics and
Anthropology. New York: Harper & Row.
Ingold, T. (1990), ‘An anthropologist looks at biology’, Man, New Series, 25.2,
208–229.
Jackendoff, R. (1983), Semantics and Cognition. Cambridge, Mass: MIT Press.
Jakobson, R. (1960), ‘Concluding statement: Linguistics and poetics’, in edited by
T. A. Sebeok (eds), Style in Language. Cambridge, MA: MIT Press, pp. 350–377.
96 N. J. Enfield
Jakobson, R. (1971), ‘Shifters, verbal categories, and the Russian verb’, in Roman
Jakobson (ed.), Selected Writings II: Word and Language. The Hague and Paris:
Mouton, pp. 130–147.
Jolly, A. (1966), ‘Lemur social behavior and primate intelligence’, Science 153,
501–506.
Kockelman, P. (2005), ‘The semiotic stance’, Semiotica 157, 233–304.
Kockelman, P. (2006a), ‘Representations of the world: Memories, perceptions, beliefs,
intentions, and plans’, Semiotica, 162, (1–4), 73–125.
Kockelman, P. (2006b), ‘Residence in the world: Affordances, instruments, actions,
roles, and identities’, Semiotica 162, 19–71.
Kockelman, P. (2007), ‘Agency: The relation between meaning, power, and knowl-
edge’, Current Anthropology, 48, (3), 375–401.
Krebs, J. R. and Dawkins, R. (1984), ‘Animal signals: Mind-reading and manipulation’,
in John R. Krebs and N. B. Davies (eds), Behavioural Ecology: An Evolutionary Approach.
London: Blackwell, pp. 380–405.
Kuper, A. (1999), Culture: The Anthropologist’s Account. Cambridge, MA: Harvard
University Press.
Lakoff, G. (1987), Women, Fire, and Dangerous Things: What Categories Reveal about the
Mind. Chicago: Chicago University Press.
Langacker, R. W. (1987), Foundations of Cognitive Grammar: Volume I, Theoretical prerequi-
sites, 1. Stanford: Stanford University Press.
Layton, R. (1997), An Introduction to Theory in Anthropology. Cambridge: Cambridge
University Press.
Lee, P. (1996), The Whorf Theory Complex: A Critical Reconstruction. Amsterdam/Philadelphia:
John Benjamins.
Lévi-Strauss, C. (1966), The Savage Mind. Chicago: University of Chicago Press.
Levinson, S. C. (2000), Presumptive Meanings: The Theory of Generalized Conversational
Implicature. Cambridge, MA and London: MIT Press.
Linton, R. (1936), The Study of Man: An Introduction. New York: Appleton-Century-
Crofts.
Mayr, E. (1964), Introduction to ‘On the origin of species: A facsimile of the First Edition’, by
Charles Darwin. (1859). Cambridge/London: Harvard University Press.
Mayr, E. (1970), Populations, Species, and Evolution. Cambridge and London: Belknap
Press of Harvard University Press.
Mayr, E. (1982), The Growth of Biological Thought: Diversity, Evolution, and Inheritance.
Cambridge, MA: Harvard University Press.
Nadel, S. F. (1957), The Theory of Social Structure. London: Cohen and West.
Newmeyer, F. J. (2003), ‘Grammar is grammar and usage is usage’, Language 79, (4),
682–707.
Norman, D. A. (1991), ‘Cognitive artifacts’, in J. M. Carroll (ed.) Designing Interaction:
Psychology at the Human-Computer Interface. Cambridge: Cambridge University Press,
17–38.
Owings, D. H. and Morton, E. S. (1998), Animal Vocal Communication: A New Approach.
Pomerantz, A. and Mandelbaum, J. (2005), ‘Conversation analytic approaches to

the relevance and uses of relationship categories in interaction’, in K. L. Fitch and
R. E. Sanders (eds), Handbook of Language and Social Interaction. Mahwah: Lawrence
Erlbaum, pp. 149–171.
Radcliffe-Brown, A. R. (1952), Structure and Function in Primitive Society. Glencoe, IL:
The Free Press.
Richerson, P. J. and Boyd, R. (2005), Not by Genes Alone: How Culture Transformed Human
Evolution. Chicago: University of Chicago Press.
Rogers, E. M. (1995), Diffusion of Innovations. New York: The Free Press.
Sacks, H. (1992), Lectures on Conversation. London: Blackwell.
Schegloff, E. A. (2006), ‘Interaction: The infrastructure for social institutions, the
natural ecological niche for language, and the arena in which culture is enacted’,
in: N. J. Enfield and S. C. Levinson (eds), Roots of Human Sociality: Culture, Cognition,
and Interaction. Oxford: Berg, pp. 70–96.
Schutz, A. (1970), On Phenomenology and Social Relations. Chicago: University of Chicago
Press.
Searle, J. R. (1969), Speech Acts: An Essay in the Philosophy of Language. Cambridge:
Sidnell, J. (2005), Talk and Practical Epistemology: The Social Life of Knowledge in a Carib-
bean Community. Amsterdam/Philadelphia, John Benjamins.
Silverstein, M. (1976), ‘Shifters, linguistic categories, and cultural description’, in
K. Basso and H. Selby (eds), Meaning in Anthropology. Albuquerque: University of
New Mexico Press, pp. 11–55.
Sussman, R. W. and Chapman, A. R. (eds) (2004), The Origins and Nature of Sociality.
New York: de Gruyter.
Tomasello, M. (1999), The Cultural Origins of Human Cognition. Cambridge, MA:
Harvard University Press.
Tomasello, M. and Call, J. (1997), Primate Cognition. New York: Oxford University
Press.
Tomasello, M., Carpenter, M., Call, J., Behne, T. and Moll, H. (2005), ‘Understanding
and sharing intentions: The origins of cultural cognition’, Behavioral and Brain
Sciences 28, (5), 664–670.
Whiten, A. and Byrne, R. W. (eds) (1997), Machiavellian Intelligence II: Extensions and
Evaluations. Cambridge: Cambridge University Press.
Wierzbicka, A. (1980), Lingua Mentalis. Sydney: Academic Press.
Wierzbicka, A. (1987), ‘Boys will be boys: “Radical semantics” versus “radical pragmat-
ics”, Language 63, (1), 95–114.
Wierzbicka, A. (1991), Cross-Cultural Pragmatics: The Semantics of Human Interaction.
Berlin/New York: Mouton de Gruyter.
Wierzbicka, A. (1992), Semantics, Culture, and Cognition. New York: Oxford University
Press.
Wittgenstein, L. (1953), Philosophical Investigations. Oxford: Basil Blackwell.
CHAPTER
6
Four-Borne Discourses: Towards
Language as a Multi-Dimensional City
Gu Yueguo
6.1 Preliminary Remarks
This chapter makes an attempt to study the use of language in terms of land-borne
situated discourse (LBSD), Written Word-borne discourse, air-borne situated dis-
course (ABSD), and Web-borne situated discourse (WBSD), with the term ‘four-
borne discourses’ as a quick umbrella term to bracket them all. Before coming to
them, it becomes paramount to make it clear what is meant by language and dis-
course. Following Chilton (2004: 16) I shall use language–L to refer to the human
capacity for language (cf. I-language in Chomsky, 1986: 22), language–l to refer to
specific languages such as Chinese, English, French, etc. The use of language–l/u
will be discourse, i.e. Chomsky’s E-language. Nothing will be said about language–L.
The focus will be on the interaction between language–l and discourse. Note that
the notion of language to be dealt with hereafter will be by default language–l unless
it is indicated otherwise. Furthermore, the Chinese language is the object of investi-
gation, and the Chinese social, political and cultural context is taken for granted.
As is well known, theories are abundant about language–l and discourse. It obvi-
ously goes beyond the scope of this chapter to review them all (Dijk, 2007 being the
most comprehensive and up to date to consult with). What is feasible is to highlight
some major conceptual frameworks that are immediately related to the content of
the chapter. However, since the distinction between four-borne discourses and the
use of them as a way to study language use are somehow idiosyncratic, I adopt a
presentation strategy as follows, hoping to achieve better clarity and a smoother
flow of ideas. First, I offer a synopsis of the main points this chapter attempts to
argue for (section 6.2). The excavating activity of an ancient tomb in Beijing is used
to demonstrate what the four-borne discourses look like in real life, thus giving the
reader an intuitive grasp of them. Then Wittgenstein’s three metaphors about lan-
guage are reviewed and examined (section 6.3). They are important because they
serve as the conceptual foundation for bridging language and the four-borne dis-
courses. Section 6.4 makes a demographic analysis of the four-borne discourses, fol-
lowed by conceptual and more empirical analyses of them in section 6.5. The
literature review will be dealt with later in section 6.6 of this chapter.
Four-Borne Discourses 99
6.2 Synopsis of the Main Points
Language is understood as an experiential phenomenon co-existent with a society/

culture. It is likened to a city, such as depicted by Wittgenstein (1997 [1953]: 8e):
Our language can be seen as an ancient city: a maze of little streets and squares, of
old and new houses, and of houses with additions from various periods; and this
surrounded by a multitude of new boroughs with straight regular streets and uni-
form houses.
In view of Chinese, the Chinese language city arguably started being built as
early as Peking Man (Homo erectus, 400,000–200,000 BP; see Fairbank, 1997:31)
who was excavated in a Beijing suburb in 1929. The written Chinese characters as
constituting a writing system could only be dated to Shang Dynasty (1765–1122
BC) with oracle bone inscriptions as its most ancient ancestor. From Peking Man
to Shang Dynasty there had been only oral-aural discourse (i.e. LSBD), or what
Ong (1982: 6) calls ‘primary orality’. The use of Chinese characters gave birth to
WWBD, an extra-dimension, as it were, on the language cityscape. The use of tele-
communications technology first in telephone calls, then radio and TV broadcast-
ing dispatches oral messages, thus creating what Ong calls ‘secondary orality’, or
ABSD in the terminology of this chapter. The ABSD in China is less than a hundred
years old, and constitutes a third dimension on the Chinese language cityscape.
WBSD in China is only a very recent event, and becomes a fourth dimension of
the existing language cityscape.
The four-borne discourses can be demonstrated with real-life occurrences. On
17 January 2000, a team of archeologists was authorized to launch an emergency
excavation of an ancient tomb which was previously found and tampered with by
grave robbers. The archeologists’ excavating activity is an instance of land-borne
situated discourse. On 18 August 2000, CCTV.com ran an online synchronous
one-way video two-way keyboard chatroom. This is an instance of Web-borne situ-
ated discourse. On 20 August 2000 CCTV1 sent its crew to the excavation site and
televised live the excavating activity for nearly two hours. The televised activity
being shown and watched on TV is an instance of air-borne situated discourse.
Figures 6.1a–6.1c show three screen shots visually illustrating the three-borne situ-
ated discourses just discussed.
In the meantime, the discovery and the excavation activity were extensively
covered in newspapers, thus generating WBSD.
6.3 Metaphors and Theorization of Language Studies
6.3.1 Wittgenstein’s Three Metaphors
As Aitchison (2003: 39) observes, the ‘usefulness of metaphor in theory building

has of course long been recognized, in a variety of scientific fields’. As a linguist,
her focus naturally by profession will be on metaphors on linguistic theorizations.
100 Gu Yueguo
(a) (b) (c)
Land-Borne Situated Discourse Web-Borne Situated Discourse Air-Borne Situated Discourse
Figure 6.1
The metaphors she has reviewed include the following: (1) language as conduit;
(2) language as tree; (3) language as waves and ripples; (4) language as game;
(5) language as chain; (6) language as plants; and (7) language as building. Since
the fourth and seventh metaphors are directly relevant to the present chapter, we
take a closer look at them respectively. Language as game was ‘originated with
Ferdinand de Saussure’ (Aitchison, 2003: 43–44). Wittgenstein, as Aitchison points
out, also adopts game metaphor, though the games Saussure and Wittgenstein had
in mind while theorizing are not quite the same. Whereas Saussure’s game is
primarily a chess game, Wittgenstein’s notion of game is a series of games with fam-
ily likeness, ranging from chess game to games such as ‘ring-a-ring-a-roses’ (see
Wittgenstein, 1997 [1953]: 5e). Moreover, Wittgenstein’s game does not need to be
a real already existing game for linguists to model on.
Imagine a language-game in which A asks and B reports the number of slabs or
blocks in a pile, or the colours and shapes of the building-stones that are stacked
in such-and-such a place. Such a report might run: ‘Five slabs’.
The purpose of Wittgenstein’s use of game metaphor goes beyond highlighting
a game’s rule-governedness (as hinted at in Aitchison, 2003: 44) that language
games share. It is more profound:
Here the term ‘language-game’ is meant to bring into prominence the fact that
the speaking of language is part of an activity, or of a form of life. (Wittgenstein, 1997
[1958]: 11e; bold and italics mine)
It is precisely by this ‘fact’ about the speaking of language being part of an activ-
ity or of a form life that Wittgenstein’s game metaphor is related to another meta-
phor of his, viz. language as an ancient city.
In Aitchison’s review, Wittgenstein’s city metaphor gets lost in what she calls
‘buildings’ metaphor. Under the subheading ‘5.7 Buildings’ (italics original, 2003: 45),
she writes: ‘The metaphor of language as a city or house possibly began with
Wittgenstein’ (2003: 45). In the passage she cited (see section 6.2 above), Wittgen-
stein obviously did not metaphorize language to house, but language to an ancient
city. Although a city is made of houses and buildings, the image of a city and the
image of a house are not quite the same. The fact that she brackets them under
one heading is perhaps due to her desire to include Charles Fillmore’s likening
‘language to a house’ (see Aitchison, 2003: 45). Besides the game and city meta-
phors, Wittgenstein also resorts to another metaphor in talking about language:
‘Language is a labyrinth of paths.’ (Wittgenstein, 1997 [1953]: xx).
There are reasons for one to argue that Wittgenstein’s two metaphors, language
as game and language as labyrinth are the metaphors used to throw light on the
language as city metaphor. In other words, the three metaphors make an inte-
grated picture of language conceived by Wittgenstein. First, the language-as-an-
ancient-city metaphor is not accidental, but rather plays a major role in Wittgenstein’s
philosophizing in general. He starts his Philosophical Investigations with what he
calls ‘primitive language’, – the term being just brought in without clearly defining
it. It has nothing to do with the term as used in the literature of early anthropologi-
cal linguistics. Wittgenstein’s primitive language is an imagined language in which
there are only two speakers, i.e. ‘a builder A and an assistant B’. This primitive
language consists only of the words ‘block’, ‘pillar’, ‘slab’, ‘beam’. As the Philoso-
phical Investigations unfolds, the primitive language gets richer with more items
added.
This is in complete accordance with the city metaphor: A city may start with a
single street with a few buildings, and gets more and more complicated as it
expands. Analogously Wittgenstein’s primitive language commences with a few
words with which a simple language-game is being played, and gets expanded
with more words added and more language-games being played. As city dwellers
live their lives in the city, so do the language speakers in the language-city play
language-games, which, as Wittgenstein emphasizes, are in essence forms of life.
6.3.2 Approaching Language as Multi-dimensional Modes of Existence
In Philosophical Investigations, language-games are given a fairly detailed treatment,

but not the language-city metaphor. As mentioned above, the language-city meta-
phor plays a role more in the way Wittgenstein philosophizes the language-games
than anything else. So if our attention is placed on the relation between language
use and language, there is not much to be found in there. In this subsection
I would argue that Wittgenstein’s language-game metaphor is instructive for the
study of language use, while his language-as-an-ancient-city metaphor indicates a
way to the study of language (recall that language here being language–l). It has to be
pointed out here that Ayer’s (Ayer, 1968) argument for a possibility of there being
a private language, which is denied by Wittgenstein, is accepted in this chapter.
In other words, it is possible for a private language to become a public one. It is
therefore possible for the first speaker to make the first use of his or her private
102 Gu Yueguo
language before the first use becomes adopted by others to make the first public
language-game.
Unlike Wittgenstein’s primitive language which is constructed and deliberately
simplified, Peking Man’s language was primitive in the sense that it had only one
mode of existence, in comparison with the language used by the present-day Beijing
men, which is argued to have four modes of existence. As we know, between Peking
Man and the oracle bone inscriptions there had been an interim period lasting
over 390,000 years. The language or the Neolithic-Bronze Chinese language (if
I may coin this term) had an existence that was kept live on moment-by-moment
basis by the language users while making their living. It got inherited from one
generation to another via face-to-face oral-aural communication. For ease of refer-
ence, the Neolithic-Bronze Chinese will be said to have an oral-aural mode of existence.
There always exists a pending danger for a language to become extinct if it has
an oral-aural mode of existence only (many have died, and many more will follow,
see Harrison, 2007). In Chinese, the oracle bone inscriptions give it another mode
of existence, i.e. the Written Word mode of existence.
Closely associated with the oral-aural mode of existence is the mode of exist-
ence mediated by telecommunications technology, i.e. telephone, radio and tradi-
tional TV. It will be referred to as the traditional telecom-mediated oral-aural mode
of existence. Finally, the latest internet technology has provided still another mode
of existence, i.e. cyber mode of existence. By the current state-of-the-art Web tech-
nology, the cyber mode of existence is closely associated with the Written Word
mode of existence.
Metaphorically speaking, the cityscape of the present-day Chinese language
consists of four modes of existence. They are being maintained and kept being
reproduced by four-borne discourses as those demonstrated in section 2 above, to
which we turn.
6.3.3 Four-Borne Discourses: The Roles They Play in the Chinese Language
The four-borne discourses, i.e. land-borne situated discourse (LBSD), written

word-borne discourse (WWBD), air-borne situated discourse (ABSD) and Web-
borne situated discourse (WBSD), are in order of historical occurrence. The roles
they play in the Chinese language are manifold.
1. They enrich the modes of existence for the Chinese language, one on top of
another.
2. They reinforce one another by recycling some of the other’s resources. For
instance, CCTV (news channel) has recently run a program called Meiti Guangchang
(ၦԧᑓഎ), the content of which is to televise newspaper cuttings from some
influential newspapers. The cityscape of the Chinese language thus is not a static
stack of discourses, but a dynamic flux with discourses flowing around.
3. The four-borne discourses can be transformed from the LBSD to the WWBD,
and/or to the ABSD, and/or the WBSD, and vice versa. Broadcasting live the
LBSD is an instance of transforming the LBSD into the ABSD and the WBSD.
To turn a WWBD into a TV series is an instance of transforming the WWBD to
the ABSD.
Finally since the four-borne discourses are largely unplanned, except the WWBD
which is occasionally subject to deliberate regulation, the language ‘cityscape’ is
hence an emergent property from the interactions of the four-borne discourses.
6.4 The Four-Borne Discourses: A Demographic Characterization
The discussion so far serves as a ground preparation for the treatment of four-
borne discourses proper. This section makes a demographic characterization of
the four before giving them a more theoretical analysis in the following section.
6.4.1 The Four-Borne Discourses and Il/literacy
The LBSD is an essential part of language users’ everyday life. Native speakers
acquire the ability of receiving and producing it without any conscious efforts
required. The WWBD, on the other hand, is a learned discourse. It takes several
years of conscious learning for a native Chinese to be able to read written Chinese
text. It then takes several more years for him or her to be able to write a coherent
composition. It is reported that by the year 2005, the population of illiterates still
remains as high as 116,000,0001. This means that over a hundred million people
have no access to the WWBD.
According to People’s Daily (Overseas Edition) of 24 December 2007, the
Mainland Chinese TV viewers reached a total sum of 1.2 billion. In other words,
1.2 billion people have access to the ABSD. It is worth noting that the ABSD popu-
lation cuts across the division between the literate and the illiterate drawn by the
WWBD. It is due to the fact that the ABSD, being primarily oral-aural, is thus par-
tially accessible to illiterates.
Finally, although the internet was made available to the general public only about
a decade ago in China, the internet surfers, according to the CNNIC Report (in
July 2008), had reached about 253,000,000, overtaking the United States in abso-
lute number for the first time. In the terminology of this chapter, over two hundred
and a half million people in China in 2008 have access to the WBSD. It is equally
worth noticing that the WBSD population cuts across the divisions drawn by the
WWBD or by the ABSD. This is due to the fact that some WWBD literates are inter-
net illiterates, and that the majority of TV viewers are internet illiterates as well.2
Figure 6.2 is a bar chart graphically recapturing the demographic differences
between the four-borne discourses.
104 Gu Yueguo
1,400,000,000
1,200,000,000
1,000,000,000
800,000,000
600,000,000
400,000,000
200,000,000
0
n ts )
io ul ad SD BS
D
lat ad re AB
pu e to W
o at le
alp it er ab
ot Ill e
=t (b
D
BS BD
L W
W
Figure 6.2 Demographic differences between the four-borne discourses.
6.4.2 The Four-Borne Discourses and Categorization of Idiolects
An idiolect, unlike a natural language, has an intrinsic feature of fixed temporality.

That is, it synchronizes with the lifespan of an individual language user. It starts
with Motherese, i.e. LBSDs in the form of primary orality. City children, attending
kindergarten at age 3, begin to get in touch with WWSDs. They study WWSDs in
real earnest at age 6 at primary school. Nowadays, well-equipped so-called key and
well-known primary schools introduce elements of the ABSD and the WBSD at
Grade 4. For instance, pupils are shown how to download images, audio clips
and how to make a multimedia PPT presentation. Children in the countryside as
well as those of migrant workers, on the other hand, are deprived of access to the
WWSD before age 6, or access to the ABSD and the WBSD at Grade 4. So in present-
day China, with regard to four-borne discourses, idiolects begin to differentiate at
age 3 with regard to the accessibility to the four-borne discourses, thanks to the
unequal provision of educational resources. It is worth noting at this point that,
although the ABSD and the WBSD transcend physical space-time, their access
does not.
In all primary, secondary and tertiary education, the LBSD and the WWSD go
hand in hand, but the LBSD becomes an instructional and interpersonal medium,
while the WWSD becomes academic, and prestigious, since idiolects excelling in
the LBSD are not appreciated in general by the educational system.
In view of an individual’s ability of performance, the four-borne discourses
can break into two general categories, receptive and productive. The individual is
receptive when she/he can experience the discourse by being able to interpret
it (not necessarily with a successful understanding of the intended message).

She/he is productive when she/he can not only interpret the message but also
produce verbal responses. With regard to the LBSD, infants undergo a receptive
stage before maturing in productive stage. One loses productivity when one is
getting too old or frail to speak, although one’s receptive ability may remain
functional.
With the receptive/productive distinction, individuals can be grouped into
seven idiolect categories: (1) pure one-borne idiolect speakers, e.g. illiterates in
the remote mountainous areas whose activity zones are confined to the areas within
walking distance; (2) enriched one-borne idiolect speakers, e.g. illiterates as just
mentioned, but with access to TV programs in local dialects or Putonghua. Note
that illiterates may develop a good receptive ability in interpreting TV programs in
their own way; (3) expanded one-borne idiolect speakers, e.g. literates who have
passed basic literacy tests. So they become receptive to some WWBDs, as well as
ABSDs; (4) two-borne idiolect speakers, i.e. speakers who are productive in both
LBSD and WWBD, e.g. secretaries, civil servants. Note that it becomes default that
they are receptive to the ABSD; (5) expanded two-borne idiolect speakers, i.e.
those two-borne idiolect speakers who are receptive to the WBSD by surfing it a
lot, e.g. editors, literary or non-literary writers; (6) three-borne idiolect speakers,
i.e. speakers who are productive in the LBSD, WWBD and WBSD, e.g. literary
or non-literary writers who have their own blogs and websites; and finally (7) four-
borne idiolect speakers, i.e. speakers who are productive in all the four-borne
discourses, e.g. TV presenters who write books, present TV programs, as well as
write blogs and maintain websites (e.g. Zhao Zhongxiang (䍉ᖴ⼹), a well-known
TV presenter).
It can be estimated (almost impossible to pinpoint accurately) that the last cate-
gory, viz. the four-borne idiolect speakers are the smallest minority, while the major-
ity is the third category, viz. expanded one-borne idiolect speakers. So the seven
idiolect categories make a arrowhead-shaped figure, as shown in Figure 6.3. (Note
the figure is more metaphorical than statistically accurate. No one can work out
exact statistic figures. It is based on the numbers as shown in Figure 6.2 above.)
6.5 The Four-Borne Discourses: Conceptual and

More Empirical Analyses
Up to this point one may get an impression that the distinction between four-
borne discourses is in essence a distinction about discourse mediation, or about
the effects mass media exert on discourse. As it will become clear later on, the
notion of discourse mediation is only one parameter (see also section 6.6 below
for literature review), a minor one among other more important ones, which are:
(1) situatedness of discourse, (2) human spatial-temporal movement, (3) social
space-time, (4) multimodality, (5) sedimentation of discourse process and (6) acces-
sibility and availability. They are to be dealt with one by one. First, the LBSD will
be used to demonstrate them, followed by the other three.
106 Gu Yueguo
Idiolect Receptive and Speaker types

categories productive types
LBSD literary or non-literary writers,

4-borne WWBD Bloggers, and TV presenters,
idiolects WBSD e.g. Zhao Zhongxiang
ABSD
LBSD literary or non-literary

3-borne writers as well as Bloggers,
WWBD
idiolects Website keepers
WBSD
ABSD (receptive only)
expanded LBSD
2-borne WWBD literary or non-literary
idiolects WBSD (receptive only) writers, editors, academics
ABSD (receptive only)
2-borne LBSD literates with formal higher

idiolects WWBD level education and
ABSD (receptive only) producing practical writings
expanded LBSD
1-borne literates with some
WWBD (receptive only)
idiolects ABSD (receptive only) formal literacy education
enriched illiterates having access to

1-borne LBSD ABSD TV programs in local
idiolects dialects or in Putonghua
pure 1-borne illiterates in the remote

idiolects LBSD mountainous areas
Figure 6.3 An arrowhead-shaped distribution pattern of idiolects.
6.5.1 Analytic Parameters
6.5.1.1 Situated discourse

The term situated discourse as type name is used in contrast with non-situated dis-
course. It is an abstraction of all those instances of language use that have the ensu-
ing characteristics. The language use is
1. situated to a particular spatial and temporal setting of a social situation (as

defined by Argyle, et al., 1981);
2. situated to the existing functions and goals of the social situation;
3. situated to the ongoing activities of the social situation;
4. situated to the inter-subjective world of the face-to-face acting participants;
5. situated to the performance contingencies of speakers;
6. situated to the immediate and remote history of the situation.
A prototypical case of situated discourse is land-borne situated discourse. The

word land means ‘a part of the surface of the earth marked off by natural or politi-
cal boundaries or the like’ (Random House Webster’s Unabridged Electronic Dictionary).
The compound land-borne is coined by the author to bracket the following features
of LBSD:
1. In the era of primary orality, people’s communicative interaction had to be

face-to-face (i.e. co-present), at least within an earshot distance;
2. The co-presence makes it obligatory that the interactants must converge physi-
cally to a particular behaviour setting;
3. The physical convergence means that the interactants must move over physical
space and time. The ‘movement in space is also movement in time’ (Hägerstrand,
1978b[1975]; requoted from Giddens, 1984: 112);
4. Since the human body is indivisible, no human beings can participate in more
than one activity at a time (cf. Hägerstrand, 1975 and 1978).
Nowadays people are of course much better off than people in the primary oral-
ity era, but the LBSD just depicted still extensively exists. It is still a much preferred
mode of discourse. In China there are some ethnic nationalities whose discourse
is still in oral-aural mode only.
The situatedness and co-presence of LBSD have four important impacts. First,
all the participants can be uniquely identified, with the body and the person two
in one. Second, interactive ethics such as being polite, face-caring, etc., comes to
the active consciousness of the participants. Third, the here-and-now enclosure is
temporarily formed, thus creating a temporary boundary between the insider (all
the participants presently involved) and the outsider (all those who are absent).
And last, the discourse process as well as its contents will evaporate except for the
memory traces and for the hard copy records if they were deliberately being made.
These four impacts, which are often taken for granted, will become problematic in
the other three-borne discourses.
6.5.1.2 Human space-time movement and activity zone

Thanks to the LBSD’s features discussed above, participants’ movement in physi-
cal space and time frames and at the same time enables the formation as well as
distribution of LBSD. In real-life terms, movement in physical space and time
means movement from one place to another. The LBSD is a web of nodes formed
by the trajectories of everyday activities of language users. The configuration of
the web is profoundly structured by the life patterns of a speech community and
the individual language user’s daily activity zone (see Gu, 2002b: 144–145 for a
case study of week-long activities of Mr X). In the author’s field study in Inner
Mongolia, for instance, it is found that there are local herdsmen whose daily activ-
ity zone is only limited to a walking distance. They tend to be a grandpa and
grandma generation. The Mum-Dad generation, on the other hand, has a broader
activity zone, thanks to their adoption of modern transportation. One of the con-
sequences the difference in activity zones brings about is that the grandpa and
108 Gu Yueguo
grandma generation maintains the Spoken Mongolian, while the Mum-Dad gene-
ration becomes bilingual, i.e. in Putonghua and the Spoken Mongolian.
6.5.1.3 Social Space-Time

The previous section has shown the coupling of human bodily movement in physi-
cal space and time with individual members’ everyday activities. This coupling pro-
vides a bridge linking bodily movement in physical space and time with Social
Space-Time. The term social space-time is used by the author to conceptualize two
interrelated phenomena. One refers to the scope of social mobility and freedom a
social system provides for its members by way of laws, decrees, regulations, control
of resources, values, etc. (this sense will be marked by the initial capital letters,
Social Space-Time). The other refers to the chances of making a living a social system
offers its members (this sense will be signalled in initial lower case, i.e. social space-
time). The concept can be demonstrated with the age-old household registration
practice, the earliest written record of which was found in West Zhou Dynasty
(1121–771 BC).3 In Ming Dynasty (1368–1644 AD), there is a law with the ensuing
article: ‘Farmers must remain within one li. Go out to labour in the morning and
return home in the evening. Whereabouts must be made known to one another.’4
Whoever wants to leave beyond one hundred li must obtain a travel certificate
(luyin 䏃ᓩ). While ‘one hundred li’ is physical space, and from morning to evening
is physical time, the fact that farmers and their behaviour are bound by the law to
this physical space-time is Social Space-Time. The farmers’ chances for making a
living within this Social Space-Time will be their social space-time.
In the twenty-first century China, the law about household registration still
remains in effect, although it is much less restrictive. Its consequences however are
quite substantial. The fact that millions of migrant workers pour into cities to make
a living shows that their social space-time is thus made much broader than the
non-migrant countryside compatriots. This does not improve their Social Space-
Time, since they cannot become registered as regular household members in their
host cities, thus being denied access to the benefits the urban citizens enjoy.
Migrant workers bring their LBSDs with them to the host cities. Their LBSDs are
automatically made inferior to the urban LBSDs, and physically evaporate the
moment they are being produced.
In theory the less constraining Social Space-Time gives social members, the
more social space-time there will be for personal development, and social mobility.
In practice, however, not all social members can make the best of the enlarged
social space-time. One of the consequences is that the generation gap becomes
wider. To continue the example of the activity zones of Inner Mongolian family
case mentioned above. The grandparents’generation’s activity zone of LBSD remains
within a walking distance in the place where they were born. The second ‘mum-
&-dad’ generation reaches out to the county’s capital. The third generation, on
the other hand, tries to get away from the place of origin as fast as they can. The
outcome is that the third generation has gradually lost the ability of speaking the
grandparents’generation’s LBSD (i.e. the Spoken Mongolian) with oral compre-
hension failing fast on its way.
6.5.1.4 Natural multimodality and total saturated experience

In the LBSD, the face-to-face, co-present interaction between participants is natu-
rally multimodal in the sense that human sense organs including the body trunk
may all be involved in giving and giving off information (see Goffman, 1963) for
one another. Given such a session of interaction it is a chunk of life being experi-
enced by participants in flesh and blood. The term Total Saturated Experience (TSE)
will be used as an umbrella term to bracket all that is being experienced by the
actors. The TSE is a spatial-temporally bounded experience in the sense that it has
a beginning and an end (remember that no face-to-face interaction lasts forever).
It is the natural multimodality and total saturated experience in the LBSD that
are being mediated and transformed in the WWBD, ABSD and WBSD. This is one
of the crucial aspects in which the present study differs from the existing literature
on mediated discourse (see section 6.6 below).
6.5.1.5 Sedimentation of discourse process

As mentioned before, the LBSD will evaporate the moment the engagement
between the participants is over, unless measures are taken in preserving it. This
leads to the issue of sedimentation of discourse process. The LBSD has two promi-
nent ways of discoursal sedimentation, i.e. the preservation of those LBSD features
that survive after the closing of a here-and-now interaction. One is that it is prima-
rily facilitated through human memory in the sense that the LBSD’s past occur-
rences, here-and-now occurrences, and forthcoming occurrences were produced,
are being produced and are to be reproduced on the basis of human memory as
the main integrative device. The other is the behaviour setting (in Barker’s sense,
Barker, 1968), e.g. a behaviour setting fixed up for teaching, for conferencing, for
selling things, etc.
In comparison with the other three-borne discourses, the LBSD is advantageous
in total saturated experience, but weak in discoursal sedimentation, whereas the
other three are weak in total saturated experience, but advantageous in discoursal
sedimentation (see 6.5.2.1 below; see also Gu, 2008 for further discussion).
6.5.1.6 Accessibility and availability

Accessibility is associated with the affordance the four-borne discourses are intrin-
sically able to provide, whereas availability refers to the chances individual mem-
bers have in participating in the four-borne discourses. The LBSD is both accessible
and available to all members of a speech community. The difference between
members is the variety of LBSDs that are made accessible and available to them.
Against these two parameters we see the nature of a double blaze sword of the
WWBD, ABSD and WBSD (see discussion below).
6.5.2 WWBD, ABSD and WBSD: A Contrastive Study
The six parameters discussed in 6.5.1 above serve as differentiators against which
the WWBD, ABSD and WBSD are to be examined.
110 Gu Yueguo
6.5.2.1 The WWBD
Situatedness of discourse: The WWBD is a prototypical case of non-situated discourse,

using those features of situatedness discussed in 6.5.1.1 and 6.5.1.2 above as
yardsticks. For a written discourse can be produced by a single writer in a solitary
room without anybody else being co-present over a long period of time. Moreover,
the writer’s unique identity can be hidden from a reader’s knowledge.
Human spatial-temporal movement : In the WWBD, the writer’s original space-time is

embodied symbolically in the WWBD. It is always separate from the reader’s space-
time. The writer does not have to enter into any real space-time relation with any
reader, nor does any reader have to enter into any real space-time relation with the
writer. In other words, readers do not have to consume movement over space or
time in order to be engaged with the writer. This is obviously convenient and cost-
effective. Having said this, thanks to the separation and discontinuation of space-
time, the decoding of WWBD can be costly. It may take years to master the code
before being able to read it.
Social space-time : The Social Space-Time as defined in Section 6.5.1.3 is largely con-
structed, maintained and reinforced through the WWBD, e.g. Chinese laws, rules,
regulations, policies, etc. are all in the form of written discourse. As mentioned in
6.4.1 above, the WWBD is a double blaze sword: Those who are good at the WWBD
gain considerable advantage and prestige, i.e. enjoying a greater social space-time,
over those who are no good at it.
Natural multimodality : The Written Word is extremely abstract, and is mono-modal

in the sense that it consists of visual static marks only. When these visual static
marks have to be used to represent what is fundamentally not visual or static, the
Written Word does not alter its own intrinsic property of being mono-modal, unlike
the ever-expanding Web technology (in this connection see Section 6.5.2.3), but
transforms the multimodal content into its visual static mode. I propose to call this
transformation mono-modalization. Many novel-based TV series, e.g. The Dream of
Red Chamber, Harry Potter, multimodalize, through actors’ acting, the mono-modal-
ized print text so as to make them available to illiterates.
Sedimentation of discourse process : In languages with orthographic writing systems,

the LBSD can be abstractly reserved in writing, e.g. through transcription, but with
heavy losses of information. For instance, its physically anchoring property is lost
or reduced to a narrative description. For another, human voice qualities, proso-
dies, facial identities, etc., are lost or reduced to a couple of adjectives (see Gu,
2008 for detailed discussion). What has been gained, however, is the Written Word’s
great advantage over human memory with regard to time. As we all know, human
memory both increments with, and decays over time. For a long time in Chinese
history the Written Word has been the dominant medium in reserving and accu-
mulating knowledge and cultural heritage. According to Bodde (1957: 10), it ‘has
been estimated, . . . that as late as 1750 more books had been printed in China –
the original home of printing – than in the rest of the world put together.’
Accessibility and availability : Unlike the LBSD, which does not require conscious
efforts for any native language speaker to learn it, the WWBD does not make itself
easily accessible to the native language user, nor did it make itself easily available to
him or her during the early days. The oracle bone inscriptions were only available
to the rulers of the land. The same was true of the later bronze inscriptions. It
became easily available only after automatic printing and cheap production of paper.
6.5.2.2 The ABSD
Situatedness of discourse : While the LBSD is always confined to a fixed location such
as a building, an office, etc., the ABSD is lifted from its land-based station to open
air, the boundary of which is only set by the power of the technology being used.
(The restrictions on the reception end set by the local authorities are a different
matter!). There are three variations of situatedness. (1) Situated discourse proper:
The activity is a LBSD, which is broadcast live, such as the excavating activity in
Section 6.2. (2) Situated to a studio: The activity is a LBSD, but in a studio, i.e. a
special workplace for radio or TV staff. (3) Blended settings and loss of situated-
ness: TV series, films and other imaginative multimedia programmes are no longer
situated to a particular behaviour setting, or to a ‘here-and-now’ gathering of co-
present participants.
Human spatial-temporal movement: In the LBSD, the body and the person are in one,
while in the projected space, the body and the projected person are separate. A
TV-projected person is one person simultaneously over two spaces, the land physi-
cal space (e.g. the projecting TV studio or where the body-person stays), and the
air space defined by the TV technology. Since the body is separate from the person
in the ABSD, the projected person potentially can be different both in physical
states and in character. (That’s why Osama Bin Laden’s TV broadcasts needs to be
verified!). One of the greatest advantages of the ABSD is that the projected per-
sons and their discourse can travel in all directions, while consuming little time.
Social space-time: Unlike the BBC which can broadcast real-life parliamentary
debates live, the role China’s CCTV plays is to disseminate what has been ‘fixed’
already. So the ABSD in China does not take part in making Social Space-Time, but
only in maintaining the status quo of the existing Social Space-Time. With regard
to social space-time (i.e. chances for making a living), however, the ABSD’s role is
drastically different. ABSD producers and the projected persons gain a consider-
able amount of social capital, thanks to the all-direction, almost timeless travel!
(See also the discussion about accessibility and availability below.)
Natural multimodality: While in the LBSD, the participants are both producers and
consumers of here-and-now discourse, it is not the case with the ABSD. The ABSD
112 Gu Yueguo
creates two general classes of participants: the producers and the consumers. The
ABSD producers’ interaction between themselves can be naturally multimodal,
but the interaction between the producer and the consumer is oral-aural only in
radio, and oral-aural plus visual in TV. Since the primary purpose of making ABSDs
is to promote interaction with consumers, the ABSD is not naturally multimodal,
but mono-modal in case of radio reception, or dual-modal (i.e. visual and aural) in
case of watching TV shows.
Sedimentation of discourse process: There are two types of ‘memories’ in the ABSD.
The producers and consumers provide human memories. All the programmes,
radio and TV alike, are ‘remembered’ in technology-mediated memory, e.g. ana-
logue cassettes, digital hard disks, etc. The great advantage of this technology-
mediated memory is that it freezes voice qualities, still and moving images against
the ruthless time erosion. It is an indispensible and irreplaceable supplement to
the Written Word records.
Accessibility and availability: The ABSD’s oral-aural and visual modality, being close
to the LBSD’s natural multimodality, makes the ABSD more accessible to Written
Word illiterates. Moreover, the mass and cheap production of radio and TV sets
makes the ABSD more readily available to the general public than ever before. In
the long run it helps improve the social space-time of the Written Word illiterates
(e.g. making them better informed than otherwise).
6.5.2.3 The WBSD
Situatedness of discourse: In the LBSD, each participant (i.e. body-person in one) is

uniquely identified, and verified by the co-present participants, while in the ABSD,
the body and the projected person, as pointed out above (see 6.2.2), are separate
and can be different in physical states and character. In the WBSD, on the other
hand, the body and the person, in the state-of-the-art technology, are not counted
as constituting a unique identity to be admitted into the networks. IP address5 and
a password are used instead of the body-person. The separation of the real body
and the real person from the Web identity recognition is total, although in net-
conferencing, video chat, etc., there is a projected person similar to that in the
ABSD.
As in the ABSD, there are variations of situatedness in the WBSD. (1) Situated
discourse proper: The activity is an LBSD being broadcast live on the Web;
(2) Synchronous situatedness: Cyber interactants, being physically located in two
or more different behaviour settings, are engaged in chatting via keyboard, or mike
or Web cam; (3) Asynchronous: There is no synchronous engagement between
cyber interactants. It is no longer being situated (i.e. in the sense as defined in this
chapter).
Human spatial-temporal movement : As pointed above, in the LBSD, any movement in

space means a proportional amount of time consumed at the same time. In the
ABSD, the projected person’s movement over the projected space (i.e. the travel-
ling time the body-person would have spent) is merged into ABSD time. (That is
the major reason why time is so expensive in TV channels!) In the WBSD, in con-
trast, space and time are merged into each other. The projected persons/messages
fly around freely in the cybernetic space-time, regardless of the physical space of
the Earth or the time zones set up by the international bodies (note that the speed
of bandwidth effecting transmission is a different matter).
Social space-time : The WBSD, like the ABSD, is currently used to disseminating
and maintaining the existing Social Space-Time rather than constructing it. On
the other hand, massive investments have been made in setting up Web-based
educational programmes with the aim of providing the remote and impoverished
areas with the best educational resources. In this regard, the WBSD can be con-
structive in improving the social space-time of those who are handicapped due to
harsh physical environments, which, as things have actually turned out, still remains
a remote ideal.
Natural multimodality: The state-of-the-art Web technology still relies heavily on

haptic modality for Human-Machine Interaction (HMI). It will not be too long
before using oral and visual modalities for HMI. Having said this, the WBSD will
not be naturally multimodal in the sense as defined in 6.5.1.4 above. A full-blown
multimodal HMI will result in a vicarious experience at best.
Sedimentation of discourse process: The Web technology in principle knows no bounds

in machine memory. Anything occurring on the Web in principle can be reserved
in memory indefinitely. What is more important is that all networked users can
exteriorize what is in their mind by publishing it online, which instantly becomes
publicly available. Unlike human memory, there is no tensed time in machine
memory. In other words, the machine does not care whether tomorrow is future,
while yesterday was past, and here-and-now is present. Its time is tenseless! The
WBSD has revolutionized beyond measure the sedimentation of human knowl-
edge and cultural heritage.
Accessibility and availability: The current Web technology, due to the limitations of
HMI discussed above, confines its accessibility to literates. Its availability is also lim-
ited to those who can afford it.
6.5.3 The Four-Borne Discourses: Maintenance, Reproduction and Creativity
The analytic differentiators about the four-borne discourses discussed above are of
course not their only distinctive properties, but they are adequate for the purpose
of this chapter. As a way of wrapping up the discussion, let us briefly return to the
relation between language–l and discourse, or in Wittgenstein’s metaphorical termi-
nology, between language as an ancient city and language use as language-games.
114 Gu Yueguo
The examination of this relation can be made by exploring this question: In what
ways do the four-borne discourses contribute to the maintenance, reproduction
and creativity of the language city?
The LBSD is a tangible immediate reality of everyday life. Most of our waking
hours are spent in talking and most of talking is part of highly routinized daily
activities. For most people, there are fixed activity zones radiating from home to
workplace or farmland, from home to groceries, from home to school, etc. It is the
routinization of activities over a fixed activity zone that help maintain the life spans
of dialects or languages – a proposition I attempted to argue for in Gu (2002b). The
LBSD unites people through the same Spoken Word, but at the same time divides
them into physical space-time bounded clusters, e.g. dialect zones, activity zones,
Korean towns recently formed in Beijing and Weihai (Shandong Province), etc.
In a way the LBSD is the most conservative among the WWBD, ABSD and WBSD in
preserving spoken dialects or languages in China.
With regard to idiolects, the LBSD brings in dynamism into idiolects, as individ-
uals are engaged in social mobility, and migration. This dynamic flux is counter-
balanced by the WWBD which normalizes (i.e. by levelling individual differences)
and standardizes (i.e. by imposing conventional practices on individuals) idiolects
through formal education.
The LBSD is historically the oldest mode of discourse. It is assumed to be the
foundation for the other modes of discourse in the sense that it generates for the
latter (1) a knowledge base, (2) patterns of human activities and (3) ethics of inter-
personal interaction. This foundation is crucial for the interpretation of the other
modes of discourse. It furthermore acts as a springboard for the imagination and
creativity of the other modes.6
As it is, the WWBD remains the most viable means for maintenance, reproduc-
tion and creativity of the Chinese language. The ABSD remains the most viable
means for maintaining a sense of immediacy across physical boundaries, which can
never be achieved in the LBSD. The ABSD’s contribution to the reproduction of
language use lies in its capacity for mass publicity and awareness-raising. The
ABSD’s contribution to creativity of language use, on the other hand, seems to be
largely dependent on the WWBD for inspiration and conceptual breeding. Writers
for TV series, e.g., start by composing in the WWBD and then translate the ortho-
graphic text into the ABSD. Time will come when the TV series writers compose
them straightway without composing in the WWBD first, like painters just draw
without putting what is in mind first into words.
The WBSD, as it is, is still rapidly evolving as the technology keeps updating its
capabilities. It started with heavy reliance on the Written Word, but soon became
capable of processing the audio input and output, and in no time proved its might
in processing video streams. It won’t be too long before it will be able to handle the
conversion from the Spoken Word to the Written Word and vice versa.7 By then,
the WBSD will redefine the traditional status of orality and literacy, which is cur-
rently based on the WWBD. The WBSD will be a melting pot absorbing both the
WWBD and the ABSD.
The biggest contribution the current WBSD is making to the language is the
circulation and mass consumption of millions of millions of unedited, almost
spontaneous idiolects via ‘BBS’ style message boards, and blogs (cf. Crystal, 2001)
which are traditionally rooted to the ever-evanescent LBSD. Admittedly it remains
to be seen how many of those idiolects will eventually be sedimented as cultural
heritage, but their sheer ever-growing volume is staggering!
6.6 Literature Review: Where Does the Present Study Stand?
In terms of academic and intellectual debts, the present study has drawn inspira-
tions primarily from the ensuing areas, discourse analysis, corpus linguistics, socio-
linguistics, media studies, human geography and sociology. It is part of a series of
studies based on the Spoken Chinese Corpus of Situated Discourse (SCCSD, see
Gu, 2002a; also www.multimodal.cn). In this connection the present study is frag-
mented, leaving many loopholes and undefended assumptions that are only being
fixed in other places. Having said this, the chapter still has a focus of its own: It
presents a conceptualization of Wittgenstein’s metaphor of language as an ancient
city in such a way that it can be empirically investigated, using the SCCSD just
mentioned.
It is true that in Wittgenstein’s own program, the language city metaphor can
also be empirically investigated, i.e. in terms of language-games. My reservation is
that his notion of language-games is too micro for the purpose of building a lan-
guage city. It is too micro in the sense that his language-games are like stones,
slabs, bricks, cars, beams, etc., i.e. raw materials, as it were, for building a city. What
is missing is a skeleton framework, a layout, for language-games to be located in
order to function properly. The skeleton framework has an important feature, i.e.,
it is not fixed once for all, but evolving, as time goes by (note that historicity is
an essential ingredient in Wittgenstein’s ancient city metaphor). The four-borne
discourses as dealt with in sections 6.4 and 6.5 make a skeleton framework for the
cityscape of the Chinese language. One and the same type of language-game can
be played in four-borne discourses through transformation and mono- or multi-
modalization (unfortunately this process cannot be dealt with in the present study,
but see Gu, 2008 for partial treatment). And the four-borne discourses can of
course cater for different language-games.
Since the present theme is about the skeleton framework, i.e. the four-borne
discourses, the literature review will be concentrated on it only, otherwise it would
take minimally a book-length job to complete it. Hence, the review will include two
immediate areas: media studies and discourse analysis. My conceptualization
of LBSD owes a great debt to human geography in general, and Hägerstrand’s
theory of time geography in particular. The review of Hägerstrand is made in
Gu (2002b).
6.6.1 Media Studies
Media studies as an academic discipline gets impetus from the development of the
modern mass media such as newspaper, telephone, radio, TV and the latest inter-
net with a great amount of literature having been accumulated. As a space-saving
116 Gu Yueguo
strategy, Danesi’s (2002) excellent synthesis of the semiotic approach will be used
as the representative of the state-of-the-art research. The present study shares
with media studies the ensuing assumptions: (1) The human world is a mediated
world; (2) The development of the mass media (print medium, telecommunica-
tions media and digital media) has exerted tremendous impacts on the social
and cultural developments of mankind. What makes the present study differ from
media studies is its exclusive concern with language. And thanks to this focus,
its conceptualization of the roles media play, and research method will differ
accordingly.
6.6.1.1 Media and a Fundamental Metaphor

Theorization in media studies is underpinned by a fundamental metaphor con-
ceived by McLuhan, one of the founding figures of the discipline, the subtitle of
whose book is the very metaphor: Understanding Media: The Extensions of Man.
McLuhan (1964: 79), following Henri Bergson, suggests that
language is a human technology that has impaired and diminished the values
of the collective unconscious. It is the extension of man in speech that enables
the intellect to detach itself from the vastly wider reality. Without language, . . .
human intelligence would have remained totally involved in the objects of
attention.
Apart from language as the very basic human technology, which can hardly be
regarded as being invented by humans, there are three major technologies genu-
inely invented by man that have been scrutinized in media studies: print, telecom-
munications and digital computing.
6.6.1.2 Semiotic approach to media studies

McLuhan’s (1964) work reads more like a learned ‘lyric’ than an academic thesis.
He is regarded by some as ‘the erudite but generally unintelligible Toronto profes-
sor’ (see Koch, 1996: 1). So his pioneering work functions more like a source for
inspiration than a model to be followed directly. Academically, media have been
studied in various ways, the most influential of which is a semiotic approach. The
primary concern of the approach is to examine the intricate ways in which a
medium represents ideas and how some ideas become deposited as cultural heri-
tage. Methodologically speaking, the examination boils down to interpretation of
signs. To quote Danesi (2002: iv):
. . . Semiotic inquiry aims to unravel the nature of the relation X = Y. The X is

something that exists materially. It could be a word, a novel, a TV programme,
or some other human artifact. The Y is what the artifact means in all its dimen-
sions (personal, social, historical). Figuring out what the meanings of Y are con-
stitutes the sum and substance of semiotic method. This procedure is generally
referred to as ‘interpretation’.
Adopting Saussure’s terminology, X is the signifier; the meaning or meanings,

Y, that it generates is the signified.
6.6.2 The Present Study: Rethinking of ‘the Extensions of Man’
At this point one may wonder if there is any point for the present study to
reinvent the wheel instead of adopting the well-established semiotics framework.
There are several reasons for this. First, McLuhan’s conceptualization of the
media’s extensions of man departs from the abstract state of ‘man without lan-
guage’, while our departure point is ‘men living a life with speech in a physical
space-time setting – i.e. the LBSD. McLuhan’s project is obviously more abstract at
conceptual level, and at the same time much broader in scope than ours. Second,
thanks to the difference in the departure point, our interest in technology is
motivated by the question: In what ways do technologies make an impact on ‘men
living a life with speech’? while McLuhan is concerned with the ways various media
extend man’s physical and mental capabilities. In our conceptualization, ‘men liv-
ing a life with speech’ is not regarded as being mediated, or rather as being natu-
rally mediated, whereas in McLuhan’s, it is already mediated. Third, instead of
conceptualizing the relation between man and media as the latter extending the
former, we conceptualize the interaction between ‘men living a life with speech’
and the various technologies as a process of transformative partnership. Take the
Written Word (i.e. a writing system) for example. The Written Word can be used
to transcribe what is audible in ‘men living a life with speech’, thus transforming
what is oral-aural into static combination of mute letters. It can be used to record
what happens, transforming a real-life activity into a narrative text. It can also be
used to create documents stating laws, regulations, etc. that will prescribe a Social
Space-Time for what men can or cannot do while living a life with speech, thus
transforming the latter into a new mode of living.
The adoption of special terminologies, i.e. LBSD, WWBD, ABSD and WBSD, is
intended to reflect our different conceptualization. The LBSD is naturally multi-
modal in experience, and naturally mediated, if we prefer to call it that. It is not
depersonalized, or dislocated or disembodied. The identities of the participants
are mutually recognized. It represents a process of TSE. When a WWBD is pro-
duced to transcribe it, or report it, or describe it, etc., the transformative process
entails both gains and losses. Similarly when an ABSD, or WBSD is produced in
connection with the LBSD, again the transformative process entails both gains and
losses. In our conceptualization, the LBSD serves as the basis as well as the refer-
ence framework for studying technologically mediated discourses.
The methodology adopted in studying the process of the transformative part-
nerships between the LBSD and the other technologically mediated discourses is
corpus-based. Samples of the four-borne discourses are collected and compiled
into corpora for systematic investigation. Since the data consist of audio, video
and orthographic texts, a fresh way of processing such data, i.e. agent-oriented
118 Gu Yueguo
modelling plus segmentation and annotation, is called for (see Gu, 2006a, 2006b,
2007a, 2007b, 2008 for details).
6.6.3 Discourse Analysis and Media
The present study covers two strands in discourse analysis. One strand is a focus on
naturally occurring discourse, e.g. conversation analysis, professional discourse,
workplace discourse, etc. The other is a focus on media discourse. In view of both
strands the present study differs from the existing literature in its alternative per-
spective in looking at more or less the same phenomena. The present study looks
at naturally occurring discourse with a focus on situatedness, human mobility and
the impacts of human geography on discourse production and reproduction.
Media discourse analysis itself has three strands. The first is a critical strand.
It examines media discourse critically (e.g. van Dijk, 1988; Fowler, 1991) by bring-
ing to light the control of dominating ideology, and political power on media
discourse production and reproduction through textual linguistic analyses. The
second is a semiotic strand (e.g. Hodge and Kress, 1988; Kress and Leeuwen, 1996),
which is closely related to the semiotic approach in the media studies mentioned
above. The third strand is derived from sociocultural analysis of mind (e.g. Wertsch,
et al., 1995; Wertsch, 1998), and Scollon (1998), Scollon and Scollon (2003) Le
Vine and Scollon (2004) are studies representing this strand. Scollon’s notion of
mediated discourse is derived from Wertsch’s concept of mediated action, which
‘involves focusing on agents and their cultural tools – the mediators of action’.
‘The task of a sociocultural approach is to explicate the relationships between
human action, on the one hand, and the cultural, institutional, and historical
contexts in which this action occurs, on the other.’ (Wertsch, 1998: 24; italics
in original). Their position is that, since language is perhaps the most important
of all cultural tools, all the use of language hence is mediated discourse, which
echoes Bergson’s position on speech/language as a human technology (see 6.
6.1.1 above).
Broadly speaking, the present study is philosophically situated to the third
strand of media discourse analysis. The fact that it does not adopt the terminology
of mediated action, or mediated discourse is due to the ensuing considerations.
First, the starting point of investigation, as mentioned in 6.6.2 above, is not the
blank state of human mind, but a human mind with a speech potential. The basic
phenomenon of investigation is a piece of situated discourse, i.e. a chunk of
here-and-now face-to-face activity occurring at a particular human geographic
environment. It is not a sentence, or a text, but ‘a form of life’ in Wittgenstein’s termino-
logy. This piece of naturally occurring land-borne situated discourse is an initial, naturally
mediated building block, as it were, for a language study. As a type of language use
looked at from a speech community as a whole, the LBSD is not mediated, it is a
naturally grown one. It is mediated only to individual members of the community,
for when an individual is born, she/he is being socialized into this default mode of
language use.
Of course, the LBSD can also be examined by using the concept of mediated
action and mediated discourse in the frameworks laid down by Wertsch and
Scollon. But the price for doing so is that the LBSD will not be so readily adopted
as a reference framework for the other three-borne discourses. The primacy of
LBSD owes a debt to Ong’s insightful study of orality (Ong, 1981 [1967], 1982; also
Scheunemann, 1996).
6.7 Final Remarks
The LBSD is a study of life activities; while the WWBD is a study of how knowledge
has been created, reserved, revised, rejected or forgotten. The ABSD and WBSD
provide bridges and fresh blood-flowing channels for the previous two. It is by all
means possible that new yet-to-be-borne technologies will come and add more lay-
ers of discourse on the existing ones. In a word, a language is both ancient and
new; it is ever evolving by growing new things (i.e. new language games), while
shedding off old ones to oblivion (i.e. putting them in museums, i.e. written books,
or cassettes, or hard disks). It is always incomplete and unfinished until human
life comes to a close. It is by no means a fantasy it will not be too long before the
four-borne discourse texts can be freely transformed into one another through
the mono-modalization and multimodalization technology. By then a print novel
will be automatically acted out not through human actors, but by human-looking
digital agents (see Gu, 2006a, 2007a, 2007b)!
Notes
1 Source: www.news.cn (ᮄढ㔥) as per 18 March 2008.

2 The internet surfers count only 19.1 per cent of the total population of 1.3 billion,
whereas the TV viewers add up to over 90 per cent. See www.cnnic.cn
3 seeǉ਼⼐·⾟ᅬ·ৌ⇥Ǌ
4 “‫ݰ‬Ϯ㗙ϡߎϔ䞠П䯈ˈᳱߎᲂܹˈ԰ᙃП䘧ⳌѦⶹ”, see ∳ゟढ, “Ё೑᠋㈡ࠊᑺⱘ
ग़৆㗗ᆳǄ䕑ǉҎষᄺϢ䅵ߦ⫳㚆Ǌ2002˄01˅
5 Identities such as screen names, login names or real names are adopted for human
access convenience. They are immaterial to the accessibility of a network.
6 It will take a full-blown paper to substantiate the position proposed here.
7 It involves speech-to-text and text-to-speech technology. Presently text-to-speech
technology works quite well under laboratory conditions.
References
Aitchison, J. (2003), ‘Metaphors, models and language change’, in H. Raymond (ed.),

Motives for Language Change. Cambridge: Cambridge University Press, pp. 39–53.
Argyle, M., Furnham, A. and Graham, J. A. (1981), Social Situations. Cambridge:
120 Gu Yueguo
Ayer, A. J., (1968), ‘Can there be a private language?’, in G. Pitcher (ed.), Wittgenstein:
The Philosophical Investigations. London: Macmillan, pp. 251–266.
Barker, R. G. (1968), Ecological Psychology: Concepts and Methods for Studying the Environ-
ment of Human Behavior. Stanford: Stanford University Press.
Bodde, D. (1957), China’s Cultural Tradition. New York: Rinehart & Company, Inc.
Chilton, P. (2004), Analysing Political Discourse. London: Routledge.
Chomsky, N. (1986), Knowledge of Language. London: Praeger.
Crystal, D. (2001), Language and the Internet. Cambridge: Cambridge University Press.
Danesi, M. (2002), Understanding Media Semiotics. London: Arnold.
Fairbank, J. K. (1997), China: A New History. London: The Belknap Press of Harvard
University Press.
Fowler, R. (1991), Language in the News. London: Routledge.
Giddens, A. (1984), The Constitution of Society. Berkeley: University of California Press.
Goffman, E. (1963), Behavior in Public Places. New York: The Free Press.
Gu, Y. (2002a), ‘Sampling and representativeness in compiling Chinese Spoken
Corpus of Situated Discourse in Beijing area’, (࣫Ҁഄऎ⦄എेᐁ䆱䇁䇁᭭ᑧⱘপ
ḋϢҷ㸼ᗻ䯂乬) 䕑Ѣǉܼ⧗࣪Ϣ21Ϫ㑾Ǌ in Globalization and the 21st Century,
Beijing: Social Sciences Publisher, pp. 484–500.
Gu, Y. (2002b), ‘Towards an understanding of workplace discourse’, in C. Candlin
(ed.) Research and Practice in Professional Discourse. The City University of Hong Kong
Press, pp. 137–185.
Gu, Yueguo, (2006a), Agent-oriented modeling language, Part 1: Modeling dynamic
behavior˄෎Ѣ㾦㡆ⱘᓎ῵䇁㿔(AML)ϔ˖ࡼᗕ㸠Ўᓎ῵˅Proceedings of the 20th
International CODATA Conference, Beijing. Published by the Information Centre,
the Chinese Academy of Social Sciences.
Gu, Y. (2006b), ‘Multimodal text analysis – A corpus linguistic approach to situated
discourse’, Text & Talk, 26–2, pp. 127–167.
Gu, Y. (2007a), Multimodal Text Based Analysis of Power and Situated Discourse in
Chinese Context — Power modeling and discourse datamining. Invited speech at
the Second Roundtable Discourse Analysis, The City University of Hong Kong.
Gu, Y. (2007b), Segmenting and annotating a multimodal corpus – with special
reference to SCCSD. Keynote speech at the 3rd International Corpus Linguistics
Conference, the University of Birmingham, UK.
Gu, Y. (2008), From real-life social situation to video stream-based datamining. Plenary
speech at the 2008 CCID Workshop on Corpus Linguistics, Beijing China.
Hägerstrand, T. (1978), ‘A note on the quality of life-times’, in Carlstein, Parkes and
Thrift (eds), pp. 214–224.
Hägerstrand, T. (1975), ‘Space, time and human conditions’, in A. Karlqvist (ed.),
Dynamic Allocation of Urban Space. Farnborough: Saxon House, pp. 3–12.
Harrison, K. D. (2007), When Languages Die. Oxford: Oxford University Press.
Hodge, R. and Kress, G. (1988), Social Semiotics. Cambridge: Polity Press.
Koch, T. (1996), The Message is the Medium. Westport: Praeger.
Kress, G. and van Leeuwen, T. (1996), Reading Images. London: Routledge.
LeVine, P. and Scollon, R. (eds) (2004), Discourse and Technology: Multimodal Discourse
Analysis. Washington, DC: Georgetown University Press.
McLuhan, M. (1964), Understanding Media. London: Routledge & Kegan Paul Limited.
Ong, W. J. (1981 [1967]), The Presence of the Word. Minneapolis: University of Minnesota
Press.
Ong, W. J. (1982), Orality and Literacy. London: Methuen.
Scheunemann, D. (ed.) (1996), Orality, Literacy, and Modern Media. Columbia: Camden
House, Inc.
Scollon, R. (1998), Mediated Discourse as Social Interaction. London: Longman.
Scollon, R. and Scollon, S. W. (2003), Discourses in Place: Language in the Material World.
London: Routledge
Van Dijk, T. A. (1988), News as Discourse. Hillsdale: Lawrence Erlbaum Associates,
Publishers.
Van Dijk, T. A. (2007), Discourse Studies, 5 Vols. Los Angeles: Sage.
Wertsch, J. V. (1998), Mind as Action. New York: Oxford University Press.
Wertsch, J. V., Del Rio, P. and Amelia A. (eds) (1995), Sociocultural Studies of Mind.
Wittgenstein, L. (1997[1953]), Philosophical Investigations. 2nd edition. Translated by
G. E. M. Anscombe. Oxford: Blackwell Publishers.
CHAPTER
7
Discourse in Organizations
and Workplaces
Britt-Louise Gunnarsson
This chapter focuses on discourse in organizations and workplaces and will discuss
a number of studies which explore the complexity and diversity of communication
in modern working life. With a theoretical orientation towards sociology, organiza-
tion/network theory and social constructivism, researchers have sketched the
macro-frames which influence discourse at various levels within organizations.
Studies of the correspondence between organizational structure and discourse will
be used as a background for presentations of analyses which try to grasp and under-
stand communication in the modern organization from different viewpoints and
with a focus on different problem areas. The theoretical and methodological per-
spectives of these analyses include textlinguistics, ethnography, critical discourse
analysis, interactional sociolinguistics and conversation analysis.
A relevant question for our understanding of modern working life is related to
the effects of technological advance and globalization. The conditions for profes-
sional discourse have been influenced by a series of changes taking place in recent
decades. Technological advances have facilitated a globalization of working life,
while lifelong learning, flexibility, mobility and diversity have come to be key val-
ues in the global economy. In organizations throughout the world, we find a wide-
spread use of technology and increased reliance on the internet for internal and
external communication. Another striking feature, made possible of course by the
technological advances, is the globalization of the business world and the job mar-
ket. Throughout the world, we find transnational companies which use English as
their corporate language and employ multilingual people who can move between
jobs, between branches and between countries. The workforce mobility and work-
place diversity which characterize a globalized economy also raise issues of multi-
lingualism and multiculturalism to the surface, both in relation to the organization
as a whole and to workplace practices. For employees this new situation has come
to entail different and higher demands for literacy and communicative skills.
The purpose of this chapter is to deepen our understanding of the discourse-
related problems facing managers and employees in the globalized economy.
In the various sections of the chapter I discuss these problems from different view-
points. First, I explore the correspondence between organizational structure and
discourse at a more general level, where I use examples from analysis of small
Discourse in Organizations and Workplaces 123
workplaces and large organizations. Secondly, I view organizational discourse from

the point of view of the top-level management and discuss how an ‘organizational
self’ is constructed by means of discourse, and the role of discourse for internal
management and external marketing in the global economy. Thirdly, I view dis-
course from a workplace perspective to distinguish two types of multilingual work-
places: those which use English as lingua franca and the multilingual workplaces
with a workforce diversity. Fourthly, I discuss workplace discourse in the ‘new work
order’ with a particular focus on the consequences for the individual employee of
technological advances and a globalized economy. Last, I will sketch some topics
for future research.
7.1 Organizational Structure and Discourse
An important macro question, which is discussed here, concerns the correspond-

ence between the organizational structure and discourse. This correspondence is
seen here as a two-way relationship. Creating texts (written, spoken, and compu-
ter-mediated) therefore forms part, and an important part, of an organization’s
work. Texts1 are influenced by the social context, thus reflecting the organization
and its social structure, values, knowledge and culture. But texts also play a part in
establishing the various social dimensions of the organization. They are not only a
product of the social situation, but in their turn also shape it. We could therefore
distinguish the basic traits of a sociolinguistic order of discourse which appear in
varied forms depending on the size and structure of the organization.
7.1.1 The Sociolinguistic Order of Discourse in a Close-Knit Group
The first study I discuss provides an example of the simple sociolinguistic order of
discourse, which we find in close-knit workplace groups. Gunnarsson (1997)
presents a study of text production in a local government office with 35 employees.
The bulk of her data was collected by means of a survey given to all employees and
in-depth interviews. This study also included an analysis of the flow of texts, i.e. all
texts emanating from the office during a fixed period were categorized in relation
to sender, addressee and purpose.
In this workplace there was an obvious connection between the social and the
communicative plane. The group structure was reflected in the communicative
structure: the hierarchical social structure was reflected in communicative patterns
relating to influence and supervision; the informal group formation was reflected
in informal joint writing efforts. The social organization at the office was reflected
in the communicative organization: texts were produced as a collaborative effort;
those involved in writing documents took a collective responsibility for the text
and there was an obvious interaction between speech and writing. Another finding
related to the interaction between social and communicative factors at the level
of content. In the interviews, the employees expressed a strong feeling of group
identity, a ‘we feeling’ that distinguished this particular group from employees in
124 Britt-Louise Gunnarsson
other local government offices. This group identity manifested itself in many ways,
also in relation to writing. The interviewees made it plain that in this office they
not only had the right to cultivate a distinctive style, they made full use of this right.
They regarded themselves as pioneers of a more informal official idiom. One
conclusion that can be drawn from this is that even small groups, such as a local
government office with 35 employees, consider that they have the right to be inno-
vative in their writing. The interviews also showed that language is an element in
the build-up of the social group identity and culture. At this workplace, there was
thus a two-way relationship between discourse and the organization, i.e. texts and
discourse reflected the social organizational patterns but also played their part in
the construction of the ‘organizational self’, in the shaping of the organization.
7.1.2 The Multilayered Structure of Discourse in Large Organizations
Looking at communication within large organizations leads, of course, to a com-

plex and multi-dimensional picture of the correspondence between organizational
structure and discourse. Large organizations function within various social frame-
works with different hierarchical structures, different values, different knowledge
and different culture, and various sociolinguistic orders of discourse overlap and
intertwine leading to a complex and multilayered structure of discourse. In order
to understand communication in large organizations we must therefore consider
the interaction between different levels: the local environment, the organization,
the relevant local sector, the national sector, the national linguistic community, the
sector worldwide.
In Gunnarsson (2004a, b), this model is used for analyses of communication in
large, mainly national, European organizations. The results discussed are taken
from analyses within the research project entitled Texts in European Writing Communi-
ties2, which studied texts and text production in four mainly national writing com-
munities – banks, structural engineering firms, university occupational medicine
departments and university departments of history – in three countries: Sweden,
Germany and Britain. One purpose of the project was contrastive, i.e. a comparison
of Swedish, German and English texts of similar kinds were made; another was soci-
olinguistic, which involved studying the relationship between national culture,
organization and texts. The study comprised interviews, collecting texts, corpus con-
struction and textual analysis. Altogether, 70 executives and employees, responsible
for or involved in information activities and the production of texts, were inter-
viewed. The team also collected samples of the types of texts produced in the envi-
ronment concerned, which were then used to form the basis of a corpus containing
text types that occurred in several environments. In total the corpus consisted of 15
different text types, e.g. annual reports, brochures, letters, press releases, reports,
staff magazines. The textual analysis was multifaceted, comprising analysis at the
cognitive, pragmatic and macrothematic levels of the texts as well as analysis of argu-
mentation, discourse markers and images and image creation in companies.
The results of the analyses of the interviews and the collected texts reveal inter-
esting differences between the sectors – between banks and structural engineering
companies as well as between occupational medicine and history departments –

and also between texts produced in different national writing communities: Germany,
Sweden and Great Britain. Another finding of interest is the homogeneity within
each organization. A fairly homogeneous picture could be drawn of each com-
pany, both based on the analyses of its texts and on interview data. To a certain
extent this homogeneity could be related to the ideas steering each organization.
In particular this became striking in relation to the banks in the study, which
were guided by different ideas about organizational structure (hierarchical or
flat organization) and marketing (strong belief in advertising or in individual
interaction, i.e. service management). Part of the variation found in text patterns
between the texts from the banks could also be related to this variation in ideas.
Though the degree of importance of the national culture and the international
community may vary from one enterprise to the other, the multilayered structure
can be assumed to be fairly similar. In national as well as international enterprises,
the simple sociolinguistic order of individuals forming a close-knit group is inter-
twined with various levels of other orders leading to a multifaceted and multilay-
ered disorder.
7.2 The Construction and Maintenance of an ‘Organizational Self’
From the point of view of the top management of large – national and interna-
tional – organizations, one problem area is related to the creation of external and
internal images of the organization which can be accepted in various local settings,
another is how to create a coherent organization working for the same visions.
Texts, spoken discourse and computer-mediated communication are very impor-
tant in the creation of an ‘organizational self’ and also for the presentation of
the organization as an attractive unit in the eyes and ears of those outside and
inside it.
7.2.1 Internal Management and Marketing
Modern management theory talks about the need for internal marketing, e.g.
as a means of creating and controlling the organizational culture (Peters and
Waterman, 1982), and most top managers are fully aware of this. Discourse is at
the heart of this internal construction of the company as a unique and attractive
entity, and most organizations attach great importance to news dissemination and
storytelling within the organization. Stories of success and failure are told and
retold in organizations, thus disseminating knowledge of the behavioural patterns
to be followed or avoided in the future (Linde, 1999). In the internet era, they are
also found on the companies’ websites. Not only stories of the company and its
history, but also of individual employees and successful customers are found on
various sub-pages (Gunnarsson, 2008). Although their explicit goals vary, these
stories all help construct an ‘organizational self’ intended not only to attract new
customers, owners and prospective employees but also to help maintain the exist-
ing organizational culture.
In modern organizations, the top managers devote a great deal of time to

creating mission statements and organizational visions (Swales and Rogers, 1995;
Isaksson, 2005). A problem, however, is that a top-down communicative strategy
does not always mean that the same message penetrates the whole organization,
and organizational visions are not infrequently interpreted by employees as unin-
telligible and insignificant. Johansson (2003) presents a case study which tries
to probe this problem. Using a combination of methods, including participant
observation, discourse analysis and interviews, she analysed the organizational
communication about strategy in a Finnish–Swedish enterprise. Communication
about the strategy followed a typical top-down model, starting at group level and
ending on department level. She found that visions formulated by top managers
met different realities constructed by managers at lower levels in the company.
Managers’ attitudes, knowledge and interpretations were important individual
factors that influenced communication about the strategy. Employees did not have
the same detailed knowledge of the strategy as the managers, nor were they
given the same opportunities to acquire it. Power structures, conflicts, individual
attitudes and perspectives contributed to the successive distortion of the top man-
agement’s visions. Johansson’s study is interesting as an attempt to grasp the whole
process.
7.2.2 External Communication and Marketing
The role of discourse is no less obvious in relation to the construction of an exter-

nally addressed image which promotes success and growth. It is obvious that many
enterprises have succeeded in spreading a uniform image of themselves which is
accepted in a variety of countries and cultures. Logos, advertisements, shop design,
stories contribute to this uniformity as well as the products as such. In some cases,
the national origin of the company has come to be an essential part of its image
and the construction of uniformity. A well-known symbol for American culture
is McDonalds with its branches all over the world. Uniformity at every level –
from the product to the design of its restaurants – characterizes this franchising
company. In many countries, however, the image of McDonalds has come to be
seen as representing the negative aspects of American culture at the same time as
the company sells its hamburgers. A less ambiguous connection between national
culture and corporate image has been created by IKEA. Swedishness has become
one element in IKEA’s trademark, and it has obviously managed to turn Swedish-
ness into an image that sells. The company’s products are given Swedish personal
names, like Bosse, Björn etc. One goal in its naming policy is to present an image
of ‘trygghet’ (familiarity and security) to the Swedish audience, and an image of a
Nordic product to the international audience. Names with the Swedish letters å, ä
and ö are therefore not avoided, but in fact deliberately used to strengthen IKEA’s
exotic image.3 The ways in which customers are addressed are also aimed at
strengthening the sense of Swedishness. In countries like Austria, e.g., employees
address customers in speech and writing using ‘du’ instead of the normal ‘Sie’.
As part of its Swedish image, in other words, IKEA wishes to introduce the more egal-
itarian and familiar language and forms of address that have been used in Sweden
since the second person pronoun reform in the 1960s (cf. Gunnarsson, 2001).
Other companies have been successful in using strategies that permit greater
cultural variation. An interesting model for text production is described in Jämtelid
(2001, 2002). The concept of ‘parallel writing’ is used for the analysis of the multi-
lingual production of text in the Electrolux corporate group, where English is the
official language (cf. Gunnarsson and Jämtelid, 1999). This investigation, which is
based on interviews and text analysis, reveals an interesting model for the balance
between the local and the global. To provide a basis for company texts, e.g. bro-
chures for vacuum cleaners, a textual base in English is devised at the main office
in Stockholm, Sweden, and circulated to the sales companies in the other coun-
tries. The textual base then provides the raw material for the various consumer
brochures produced in different languages and intended for different cultures.
As Jämtelid’s analyses show, the different sales companies have chosen to incor-
porate different ideas from the textual base circulated by head office. She also
found marked differences in the arguments for the product presented in the bro-
chures, the choice of illustrations and the styles used. There were thus clear differ-
ences between Electrolux brochures written in different countries about the same
vacuum cleaner and based on the same original. The differences found between
the texts of the brochure in the various countries could indicate culturally derived
differences (Jämtelid, 2002).
This text production strategy, which was established by the Electrolux group in
the 1990s for the production of printed brochures, is interesting as it combines the
need for top-down control and group unity with the need for local and national
variation. In addition, it allowes for both linguistic and cultural variety.
In the modern technological business world, computer-mediated texts are, to a
large extent, replacing printed texts for external purposes. Products are marketed
on the companies’ websites and emails are replacing letters on paper. One impor-
tant difference between texts on a company’s website and printed documents is
their accessibility. When a company places a text on its website, it has to count on
multiple readerships. Global accessibility is a reality for texts on websites. The spec-
ified reader of the printed document is replaced by a potentially manifold reader-
ship: shareholders, staff, journalists, politicians, former and prospective customers
in countries all over the world can view the company website. Although the com-
pany can, in theory, count on multiple readerships for the texts on its website, it
does not mean that the actual number of readers increases, nor any rise in the
number of individuals who feel themselves directly addressed by the company’s
texts. The language used on a sub-page includes or excludes reader groups as does
the cultural perspective and focus of the actual text.
One consequence of the increased reliance on the internet is that company
policy on linguistic and cultural issues becomes salient and visible for different
reader groups. From a critical analysis viewpoint, the company’s textual practices
on its websites reveal, among other things, its interpretation of the concept of
diversity – a key term in the modern business world. The languages used on the
company’s customer-oriented pages reveal the groups of readers to which they

wish to sell their products, and also those that they make no effort to approach.
Are the English speaking elite among the intended readers? Do the majority
language speakers in a country belong to the intended readers? Are immigrants
and other minority language speakers in a country also addressed in a language
they fully understand?
Gunnarsson (2006) presents a study of the language practices of the websites
maintained by five transnational companies: ABB, Astra Zeneca, Electrolux, Ericsson
and Scania. All these companies claim to be world leaders in their fields: ABB is a
leader in power and automation technologies that enable utilities and industrial
customers to improve performance while reducing environmental impact. Astra
Zeneca is one of the world’s leading pharmaceutical companies, Electrolux is the
world’s largest producer of appliances and equipment for kitchens, cleaning and
outdoor use, Ericsson is the biggest supplier of mobile telephone systems in the
world and supports all major standards for wireless communication, and Scania is
one of the world’s leading manufacturers of heavy trucks and buses.
In this study, which was carried out in September 2005, the focus was on the
customer-oriented websites, i.e. the companies’.com–sites. All the companies have a
.com-site in English, functioning as a start page. For all five companies, we also find
the option of selecting a country on the start page. The number of country links
varies among the five companies, which of course reflects the global reach of each
group: ABB has 112 country links on its webpage, Astra Zeneca 105, Electrolux 67,
Ericsson 111 and Scania 39 links. The interest here relates to the language prac-
tices of the country websites. In which language(s) do the five companies present
themselves and their products to customers in different regions of the world.
Nine of the country links, namely those for Sweden, Finland, Norway, Germany,
Switzerland, France, Spain, Brazil and Japan, were analysed.
Among other things, this analysis reveals that all five companies use the main
language of the country to address customers, e.g. Norwegian on the website for
Norway, Portuguese on the website of Brazil and German on the site for Germany.
Nevertheless, there is a noteworthy difference between companies which address
their customers only in the main local language, and those that use both that
language and English. There is also a difference among the companies that offer
language options for countries with several official languages, e.g. both Finnish
and Swedish in Finland, and both German, French, Italian and Rhaeto-Romance
in Switzerland. However, only one of the companies, Scania, offers a choice between
Finnish and Swedish for Finland, and only three, Scania, Electrolux and Astra
Zeneca, a choice of both German and French for Switzerland.
There are sub-pages in English on all of Ericsson’s country websites. All the ABB
websites use the main national language, but for five countries, namely Sweden,
Finland, Switzerland, Spain and Brazil, there is a language-choice button offering
English as a second option. On the country websites of Astra Zeneca, Electrolux
and Scania, on the other hand, we only find the national language(s). On two of
Scania’s sites there is a language-choice button that offers two national languages,
and on one of the sites for Electrolux and AstraZeneca there is a similar button.
None of the companies, however, provide any information in any immigrant

language. As we know, there are large groups in many of these countries, who have
a different mother tongue from the main national language. If we take Sweden as
an example of the current language situation, statistics show that 11 per cent of
the Swedish population and 22 per cent of the workforce (persons aged between
18 and 64) were not born in Sweden (cf. Gunnarsson, 2005). These figures might
be even larger for other countries.
In theory, the internet offers global accessibility. In order to reach out to
different reader groups, however, the company has to consider regional variation.
The language used on a sub-page includes or excludes reader groups as does the
cultural perspective chosen for the text. The balance between local and global
concerns is thus related to policy and practice on linguistic and cultural matters.
For the five companies studied, accessibility seems to be reserved for those pro-
spective customers who either speak the main language of the country or have
mastered English. The companies’ claims to be global and international and to
respect the value of diversity do not entail an interest in the mother tongues of
the minorities within the different countries.
7.3 The Multilingual Workplace
The correspondence between organizational structure and discourse (cf. section 7.1)
gets even more complex in large, global organizations where language and culture
also become socially relevant issues. Although the multilingual workplace is not
a new phenomenon, the growing workforce mobility in today’s economy entails
workplace diversity of a varied and somewhat new kind. Issues of multilingualism
and multiculturalism are brought up to the surface, as are issues of dominance and
marginalization. In this section of the chapter, I dwell on problem areas related to
two kinds of multilingual workplaces. First, I will discuss studies of multilingualism
in large corporations which use English as their lingua franca. Secondly, I will dis-
cuss studies of workplaces which focus on the diversity issue at a more local level.
7.3.1 Multilingual Workplaces with English as Their Lingua Franca
Globalized economy means that transnational organizations, i.e. organizations

which operate in countries with different languages, need to choose one language
as their lingua franca. Not seldom, at least in Europe, English is chosen as the lin-
gua franca of these organizations. In countries where English is not the mother
tongue, this creates different communicative problems, in relation to both outgo-
ing and internal communication.
A number of recent studies have analysed communication with English as the
lingua franca in large, European, organizations and also focused on the various
problems related to this practice4. I will here discuss some studies which view
the problem from a Scandinavian perspective. For all the Scandinavian countries,
English has come to be the natural choice of a lingua franca at an international
level: Swedish, Danish, Norwegian and Finnish are not understood outside
Scandinavia, which indeed necessitates a lingua franca in external communication.
In the past decade, English has also come to be used as a lingua franca in
Nordic mergers, e.g. in the Finnish–Swedish company Stora-Enso. Kankaanranta
(2005a: 42) describes the communicative practices within this company in the
following way:
In practice, this language choice means that corporate-level documentation

and all reporting is done in English, and communication between different
units is mostly in English. This type of communication can be characterized as
internal communication as it is taking place ‘in-house’ in contrast to communi-
cation between buyers and sellers in which foreign languages have been tradi-
tionally needed in international business. For an individual writer of an email
message, the choice is a pragmatic one: anytime there are recipients whose
mother tongue is not that of the writer’s, the message is in English. This means
that a Swede (or a Finn) will receive an English message from another Swede
(or a Finn) if the list of recipients includes non-Swedish speakers.
The use of English as a corporate language has also impact on the recruitment of
the workforce. Gunnarsson (2009b) presents an analysis of job advertisements on
the career-oriented web pages of five transnational enterprises: ABB, AstraZeneca,
Electrolux, Ericsson and Scania. All these enterprises are large employers in
Sweden, and her study focused on advertisements for jobs in Sweden. She studied
the Swedish website of these enterprises, i.e. www.abb.se, www. astrazeneca.se etc.
For each job advertisement, she noted (1) in which language the advertisement
was written, and (2) what languages were explicitly mentioned as qualifications for
the job: Swedish, English or any other language(s). This study covered all jobs
advertised on these websites in September 2005.
As regards the first issue, 63 per cent of the advertisements were written in
Swedish, 32 per cent in English and 5 per cent in both Swedish and English.
Broken down by company, the analysis shows that all advertisements for Ericsson
are in English, while for Scania they are all in Swedish. ABB’s adverts, too, are
mainly in Swedish, while AstraZeneca has roughly similar proportions of Swedish
and English advertisements. Electrolux also uses both languages to advertise jobs
at Swedish workplaces. In fact, there are five advertisements that mix Swedish and
English, i.e some parts of the text are in Swedish and some in English.
As regards the second issue, that is what language(s) are explicitly mentioned as
requirements for the job, the results could be summarized in the following way:
For the five transnational companies multilingualism means bilingualism, i.e. a
knowledge of English and Swedish. Only 10 per cent of the advertisements men-
tion languages other than Swedish and English as a qualification. English, on the
other hand, is explicitly mentioned in 70 per cent of the advertisements. Swedish
figures less prominently as a qualification, and is explicitly required in only 34 per
cent of the adverts.
Of course, the variation in terms of the qualifications required depends on

the nature of the job. At a general level, however, we can say that multilingualism
is mainly related to knowing English and to a lesser extent the main national
language, in this case Swedish. The various other languages spoken in Sweden,
e.g. Finnish, Persian, Arabic, are not mentioned at all. Of the immigrant languages,
only Spanish, Russian and German are mentioned in the advertisements
analysed.
From these descriptions of the language situation in large organizations in
Sweden, I will turn to some studies which focus on the various problems related to
the use of English as a lingua franca. When Jämtelid interviewed people involved
in writing at different levels of the Swedish Electrolux group, she found that they
themselves referred to their corporate English as ‘bad English’ (2002: 44). A wide-
spread belief was that native speakers of English found their corporate English
poor. The director of corporate communications, however, makes it clear that ‘bad
English’ is not accepted in texts aimed at a wider readership. ‘Although a presenta-
tion brochure obviously has to be correct, it still has to be possible to send a mes-
sage or a letter or minutes to one another without them being a hundred per cent
correct’. He also elaborates on the necessity for Scandinavians to speak up even if
their English is bad:
Then of course the corporate language is quite clearly ‘bad English’. Otherwise,
you easily end up with only Americans, Brits and well-educated Swedes talking
at meetings, while Germans and Italians remain silent. What happens, rather,
is that we deliberately avoid difficult words, and that people shouldn’t feel
ashamed at all if they use bad grammar in an internal memo in English. (My
translation.)
Another consequence relates to a levelling of cultural differences in favour of a

more homogeneous style. In ongoing studies of Nordic mergers, the cultural issue
has attracted attention. In her analysis of emails written in English by Swedish and
Finnish employees of Stora Enso, Kankaanranta5 also looked at possible cultural
differences. She was able to point to certain differences between the Swedish and
Finnish writers:
In spite of the macro-level similarities, the messages written by Finns and Swedes
also exhibited differences on the level of the moves, that is, the realization of
the requests. Finns seem to favour direct requests in their writing, while Swedes
use more indirect alternatives, thus supporting the notion of the more direct
Finnish communication. (Kankaanranta, 2005a: 54.)
Nevertheless, her main conclusion relates to homogenization, in that her

study suggests that ‘lingua franca interactions are characterized by a high degree
of cooperativeness and a consensual style; together the communicators aim at
smoothness, and together they construct the situational meanings’ (55).
Other studies have been more negative about the levelling of cultural differ-
ences in international meetings. Fant (1992) on the basis of his analyses of cross-
cultural negotiations says that
If systematic longitudinal research had been carried out over the past, say, three
or four decades in order to investigate the evolution of national patterns of
doing business within what is commonly being referred to as the western world,
the results would probably have indicated that, in the first place, things have
changed a great deal globally, and, secondly, that national differences have
diminished, yielding place to some sort of ‘Americanized’ style, which has served
as a model to, and has reshaped, to a greater or lesser extent the local patterns.
(Fant, 1992: 125.)
The same applies to Börestam (2005), who has analysed meetings within Nordea,
another Nordic company with English as its official language. Her study covered
meetings held in different countries (Denmark, Finland, Norway and Sweden)
and in different languages. Her conclusion is that the levelling of cultural differ-
ences which for instance relate to the overall structure of the meeting discourse
tends to favour a more Americanized style (71). Fant’s and Börestam’s conclusions
can thus be seen as pointing to a more general problem involving the consequences
for small languages and small national discourse communities of the dominance
of English and the American culture (cf. Gunnarsson, 2001).
A third problem relates to the consequences of the divide between those with
mastery of English and those without. In many organizations, the local language
dominates the spoken discourse, while the more official documents are written in
English. Language knowledge and language deficiencies are likely to create new
hierarchies and social orders. New communicative problems become salient: Who
can communicate with the top management? Who understands which texts? Who
can perform which job within the organization? Who can communicate with whom
in the workplace?
Johansson (2003) presents a case study of communication on corporate strategy
within a transnational company with its head office in Sweden. She recorded and
analysed the information flow relating to a strategy document produced at the
top level of the company. The document was written in English and formulated by
senior managers. Managers at lower levels were then supposed to present the con-
tents to the employees. Johansson interviewed top and middle managers and also
employees and recorded meetings at which the strategy document was presented
and discussed. Her conclusion is that the visions formulated by top managers
were transformed by the middle managers in accordance with their attitudes and
conceptions of reality. The employees who acquired their information via the
middle managers did not have the same detailed knowledge of the strategy as the
managers, nor were they given the same opportunities to obtain it. Johansson’s
study shows, among other things, that the use of English for important internal
documents might increase the distance between senior management and ordinary
employees.
In many cases, the divide between those with mastery of English and those with-
out increases the gap between native and non-native speakers. The choice of Eng-
lish as a lingua franca in the Scandinavian context is quite natural from the
perspective of those who grow up there. Everyone attending school in Sweden, for
instance, is taught English as their first foreign language. People growing up in
Sweden have also been exposed to English every day, as programmes on television
(news, movies, comedies etc.) use Swedish subtitles and the original (mostly
English) sound track. English is also used in much of the pop music produced and
listened to in Sweden. The increased use of the internet has also meant that many
Swedes daily read and write English for various purposes. A new type of ‘elite bilin-
gualism’ is thus gradually developing in Sweden among young people and edu-
cated adults. For individuals who have moved to Sweden as adults, however, English
is not always as easy as for those born in Sweden. Many immigrants have received
their basic schooling (not uncommonly a very short one) in countries in the
Middle East, Africa, Asia, South America and the former Eastern Europe, which
means that they might not have been taught English at all, or at least not as their
first foreign language. For them, the use of English as a lingua franca poses greater
problems than for people who have received their basic education in Sweden. The
foreign language they have to learn first, when moving to Sweden, is of course also
Swedish and not English.
In organizations which use English as their official language these individuals’
opportunities to advance are dependent on their acquisition of not only one
foreign language, Swedish, but also English. Based on interviews with immigrant
blue-collar and office workers employed in a Swedish transnational company6,
Nelson gives this picture of the need for language knowledge at different levels:
To be able to get an office job with this company, you do not necessarily need to
know that much Swedish, but you do need to at least understand English. On
the factory floor, on the other hand, you would probably not only become lonely
but also have problems if you did not speak Swedish. (Andersson and Nelson,
2005: 34.)
7.3.2 Multilingual Workplaces with Workforce Diversity
All over the world, there are workplaces where some employees, mostly minorities,
immigrants or unskilled guest workers, have to use their second, third or fourth
language at work. In many organizations, knowledge of the dominant local lan-
guage is necessary not only for advancement and a career but also for social inte-
gration into the working group. Humour, jokes, stories, anecdotes form part of the
workplace discourse and contribute to the establishment of friendship and collegi-
ality at work and also to the avoidance of unnecessary conflicts. Social workplace
patterns related to friendship, power and dominance often reflect linguistic and
communicative skills7.
An ongoing Swedish research project focuses on the day-to-day communicative
situation of immigrants employed in two different working environments: a major
Swedish company and a large hospital. By means of interviews, observations and

analyses of spoken and written discourse, the project team is studying how the
immigrant’s professional identity is constructed in the interaction with his/her
fellow workers, and how discourse makes her/him an integrated part of the work
team. With a methodological basis in interactional sociolinguistics, the attempt
is to grasp the discourse strategies used by the immigrant and her/his colleagues
to overcome communication problems and to establish a good working climate.
A striking result is the communicative awareness of the immigrants; they are aware
of their strength and weaknesses and have found ways to handle the various situa-
tions that occur during a workday. In the interviews they mention humour as a
problem at work. Interestingly enough, however, the recordings show their ability
to understand humour and also to tell jokes and stories themselves. Among other
things, these studies reveal the collaborative character of workplace discourse.
Language problems are overcome and humour and jokes are constructed in col-
laboration between the second language speakers and colleagues speaking their
mother tongue.8
Though the link between workplace practice and organizational policy is not
simple, the individual’s social integration or marginalization could be assumed to
be related to more structural issues, e.g. what is made central in various types of
organizational discourse. From a critical perspective, it thus becomes relevant to
analyse how the organization pictures its ‘ideal’ employees: Who is chosen to rep-
resent the organization? Who is chosen as a role model for new employees?
Gunnarsson (2008) focuses on this issue by means of an analysis of transnational
companies’ career-oriented web pages. All five companies studied – ABB, Astra
Zeneca, Electrolux, Ericsson and Scania – have sub-pages presenting employees
that are linked to their situations vacant pages. By means of this analysis, she can
distinguish two groups among the companies. First, the international exchange
made possible by their employment strategies is highlighted by three of the com-
panies, namely ABB, Electrolux and Ericsson. ABB features ten Swedes who are or
have been working abroad, while Electrolux, through its success stories, constructs
a picture of a company that gives its employees, who should know several lan-
guages, an opportunity to work in many countries and meet people from other
cultures. Ericsson, too, highlights the opportunity for international exchange, i.e.
the company offers its employees the possibility of working abroad. In contrast to
Electrolux, however, it makes no mention of employees’ language skills.
Second, two of the companies, Astra Zeneca and Scania, highlight the diversity
of their Swedish workplaces. Of the 11 employees featured on AstraZeneca’s web
pages, one is an immigrant to Sweden, born in Lebanon, and of the 14 featured on
Scania’s site, one is an immigrant to Sweden, born in Iran, and one a visiting stu-
dent from Germany. Both these companies could be said to give a fairly good pic-
ture of the current Swedish job market. As regards Scania, it should be noted that
the German master’s degree student is said to be fluent in Swedish. The company’s
monolingual Swedish culture is not therefore affected by the presence of this
German girl.
Though all five companies explicitly stress various societal values, e.g. ‘diversity’,
on their websites, the analysis of the ‘success stories’ on their career-oriented web
pages show a division between companies for which diversity mainly relates to
global mobility (employees taking on jobs in different countries) on the one hand
and others for which diversity relates to the local workplace (employees working
in Sweden come from different regions of the world).
7.4 Workplace Discourse in the ‘New Work Order’
The conditions for professional discourse have been influenced by a series of

changes taking place in recent decades. Technological advances have coincided
with a globalization of working life and lifelong learning, flexibility, mobility and
diversity have come to be key values in the global economy.
The purpose of this section is to sum up research which has analysed workplace
discourse in technological organizations and workplaces. I will also speculate
about what consequences the ‘new work order’ might have for the individual
employee.
7.4.1 Technological Advances and Workplace Discourse
In organizations throughout the world, we find a widespread use of technology

and an increased reliance on the internet for internal and external communication.
Text and speech in traditional forms are intertwined with computer-mediated
communication, phone calls and video-conferencing in a way that gives presence
and simultaneity a new significance. Distance communication is increasing and
reducing the importance of being at a certain place at a certain time. The extended
access to computers and advanced technology at work have also led to an increased
role for multimodality; words and visual elements are interwoven in most texts
and professional talks are often given with both textual and visual support.
Every strand of workplace communication has, in one way or the other, been
transformed by technology. Writing at work has been affected by technology, and
new written genres and new writing processes have developed as a result of the use
of fax, email and the World Wide Web9. The use of speech in the workplace has
also been influenced by the technological advances10, and phone calls and video-
conferencing are intertwined with computer-mediated interaction leading to new
interpretations of presence and simultaneity at work.
Multimodality is not a new phenomenon11, but technological developments in
recent decades have prompted growing interest in multimodality as a phenome-
non, its function and impact on discourse. Text, speech, graphs, recorded sound
and movies are interwoven in today’s communication in a way which was not previ-
ously possible12.
The technological advances have also given rise to different types of jobs
and new workplaces13. Call centres and help desks may function in remote parts
of the world, blurring the traditional concept of workplace and organizational
connection. New genres and new communicative processes have developed at
these workplaces14.
Though these studies offer glimpses of the interlacement of discourse and tech-
nology at work and sketch methods to enable analysis, the next step should be to
analyse the consequences for the individual employees of the various changes in
the conditions for work which are found in the new, global economy.
7.4.2 The Individual Employee in the ‘New Work Order’
The term ‘new work order’ has sometimes been used to refer to the redistribution
of roles, skills and knowledge in the modern global economy, (cf. Gee et al., 1996;
Hull, 1997). The modern organization constantly needs to modify products and
customize them in order to survive in the over-competitive global marketplace. For
the modern organization, a decentralized and flexible structure with temporary
and rapidly changing networks is more valuable than stability and long-lasting
structures. Organizations have fewer levels, which means that there are fewer mid-
dle managers and less administrative staff, while at the same time the unskilled
jobs/tasks have disappeared. In terms of division of labour, this development has
meant a downward shift of responsibility to the individual employees. The organi-
zations must find means to empower their workforces, for instance core visions
and cultures should be shared by managers and workers alike. For the employees,
this new situation means different and higher demands on literacy and communi-
cative skills.
Organizations competing on the global market need to be flexible and able to
learn also in relation to discourse. Compared to earlier organizations, the modern
one is characterized by more meetings, more documentation and also an increased
need for training and advice. The ‘ideal’ employee is a flexible person who is con-
tinuously learning as she/he moves within the organization and between organiza-
tions, thus taking on new jobs and performing new tasks. Instead of unskilled
workers performing routine tasks, organizations need staff who can act independ-
ently, plan their own work and take responsibility for their role in production.
They should be able to communicate more or less directly with the top levels within
the organization. The new work order is thus demanding greater flexibility and
responsibility from individual employees. Organizational structures based on
workforce mobility and workplace diversity also make higher demands of the lan-
guage knowledge and cultural openness of the individual workers.
7.5 Topics for Future Research
As has been discussed in this chapter, professional discourse has been influenced
by a series of changes because of new technology and globalization. The ‘new work
order’ has had effects on working life in general and on the working conditions for
individuals employed in or networking with large or small organizations. A goal
for future research into discourse in organizations and workplaces is to analyse
the various ‘new’ problem areas as well as to try to sketch how difficulties could
be overcome. In this last section I dwell on a few topics which relate to the various
issues discussed earlier in this chapter.
Organizations in the ‘new work order’ are described as having fewer levels.
Hierarchies are reduced and the structural connection between organization and
workplace is blurred. Nevertheless we can assume that organizations will survive as
uniform entities. One topic for future research would be to explore the new roles
of discourse in the construction of an ‘organizational self’ in the global economy
and for the maintenance of a social order in blurred contextual frames. How are
large organizations constructed and maintained by means of the internet and
distance-communication? What role do different types of discourse play for the
formation of a uniform organization?
An important set of topics relate to the employees’ group affiliations in the new
work order. Earlier studies have found that humour, storytelling and teasing are
essential elements in the discourse of a close-knit workplace. In order to be a full
member of a working group, you must, for instance, know what jokes are socially
accepted and whom you can tease and in what way. A topic for future research is to
analyse if similar socializing patterns are also established in working groups held
together by means of distance-communication. What roles, if any, do humour
and storytelling play in such communication? Do these elements play a role also
in organizations which use English – or another language – as a lingua franca? Do
employees who sit at home working on their computers feel integrated into a work-
place group?
Fewer levels and reduced hierarchies result in vagueness and tension for indi-
vidual employees. The need for workforce flexibility also leads to a form of grow-
ing de-professionalization. Role relationships and identities have to be renegotiated
and new ones formed over and over again. For the individual this situation is likely
to create uncertainty and stress. A topic for future research would be to analyse the
role of discourse in the (re)negotiation of job identities and job relationships
within flat organizations.
Another topic which would be worth looking into relates to the role of discourse
for bottom-up influence. The physical distance between one workplace and another
is a reality in large, global organizations, which indeed means that both top-down
and bottom-up communication have to be less direct. A research question is
whether this physical distance hinders bottom-up influence on the organization.
By what type of discourse is democracy established – and hindered – in the ‘new
work order’? How does the new worker get his/her voice through to the top man-
agement level? How does top management try to steer the shop floor workers?
What role does language difference play for patterns of influence?
The new work order is also characterized by extended networking. Large com-
panies merge, but there is a framework of small group collaboration in the big
concerns. An interesting topic related to power and dominance would be to ana-
lyse if and how small network partners manage to influence the decisions of large
organizations?
In the global economy, organizations have to strike a balance between local and
global concerns as well as between economic concerns and social-societal values in order
to be competitive and trustworthy. The challenge for large organizations is to find
a balance in policy and practice between these various considerations. Important
topics for future research would be to analyse how discourse in organizations
and workplaces reflect this balance. How is diversity constructed in workplace

discourse? Who is marginalized and who is made central through workplace
discourse? What societal values are expressed in official documents and more
informal discourse within organizations?
Notes
1 In this chapter, I sometimes use ‘text’ for both spoken and written discourse.
2 A number of studies on discourse in organizations and workplaces have been
carried out at Uppsala University in Sweden. For presentations of the major ones,
see Gunnarsson (2009a)
3 Interview with a retail display designer working for IKEA (Språket, SR1, 6 September
2005).
4 Compare, for instance Nickerson (2000).
5 A full description of this study is presented in Kankaanranta (2005b).
6 In the Uppsala research project entitled The Communicative Situation of Immigrants at
Swedish Workplaces, the initial step comprised semi-structured interviews with female
and male immigrants working in a hospital and a Swedish major company.
7 A number of studies on workplace discourse have been carried out in New Zealand
within the Wellington Language in the Workplace Project. Among the large num-
ber of publications emanating from this group, I here wish to mention: Holmes
et al. (1999), Holmes (2000), Holmes and Stubbe (2003), Marra and Holmes
(2004), Holmes (2005).
8 Preliminary results from this research project, entitled The Communicative Situation
of Immigrants at Swedish Workplaces, are discussed in Andersson and Nelson (2005),
Nelson and Andersson (2005), Andersson (2006) and Nelson (2007). A full pre-
sentation of the case studies of second language speakers at a Swedish hospital is
found in Andersson (2009), which is her Ph.D. thesis. Nelson’s study of interaction
at a Swedish major company will be presented in full in her forthcoming Ph.D.
thesis from Uppsala University.
9 See, for instance, Louhiala-Salminen (1995), Bargiela-Chiappini and Nickerson
(1999), Nickerson (2000), Luzon (2002), Kankaanranta (2005b).
10 See Pan et al. (2002).
11 Cf. John Swales’ analysis of a university building (1998).
12 See, for instance, Lemke (1999), LeVine and Scollon (2004), Norris and Jones
(2005), Karlsson (2006 and 2007).
13 Interesting analyses are found in Sarangi and Roberts, (eds) (1999).
14 See, for instance, Culver et al. (1997), Landqvist (2001), Qvortrup (2002), Kong
(2002), Oliveira (2004), Silva et al. (2004), Wiberg (2005).
References
Andersson, H. (2006), Interkulturell kommunikation inom sjukvården. Den kommunikativa

situationen för invandrare på svenska arbetsplatser 1. Intervjuer. (TeFa nr 45). Uppsala.
Andersson H. (2009), Interkulturell kommunikation på ett svenskt sjukhus. Fallstudier av

andraspråkstalare i arbetslivet [Intercultural Communication at a Swedish Hospital.
Case Studies of Second Language Speakers in a Workplace]. (Skrifter utgivna
av Institutionen för nordiska språk vid Uppsala universitet 77.) Uppsala: Swedish
Science Press.
Andersson, H. and Nelson, M. (2005), ‘Communication at work. The communicative
situation of immigrants at Swedish workplaces’, in B.-L. Gunnarsson (ed.), The
Immigrant and the Workplace. (TeFa nr 41) Uppsala, pp. 27–46.
Bargiela-Chiappini, F. and Nickerson, C. (eds) (1999), Writing Business: Genres, Media
and Discourses. New York: Longman.
Börestam, U. (2005), ‘Culture and culture again. An inter-Nordic merger in the bank-
ing sector’, in B.-L. Gunnarsson (ed.), Communication in the Workplace. (TeFa nr 42),
Uppsala, 60–77.
Culver, J. D., Gerr, F. and Frumkin, H. (1997), ‘Medical information on the internet: A study
of an electronic bulletin board’, Journal of General Internal Medicine, 12, (8), 466–470.
Fant, L. (1992), ‘Scandinavians and Spaniards in negotiation’, in A. Sjögren and
L. Janson (eds), Culture and Management. In the Field of Ethnology and Business
Administration. Botkyrka, pp. 125–153.
Gee, J., Hull, G. and Lankshear, C. (1996), The New Work Order. London: Allen and Unwin.
Gunnarsson, B.-L. (1997), ‘The writing process from a sociolinguistic viewpoint’, Written
Communication, 14, (2), 139–188.
Gunnarsson, B.-L. (2001), ‘Swedish tomorrow – a product of the linguistic dominance
of English?’, in S. Boyd and L. Huss (eds), Managing Multilingualism in a European
Nation-state. Challenges for Swedish. Clevedon: Multilingual Matters Ltd, pp. 51–69.
Gunnarsson, B.-L. (2004a), ‘The multilayered structure of enterprise discourse’, Infor-
mation Design Journal + Document Design, 12, (1), 36–48.
Gunnarsson, B.-L. (2004b), ‘Orders and disorders of enterprise discourse’, in C. Gouveia,
C. Silvestre, and L. Azuaga (eds), Discourse, Communication and the Enterprise. Linguistic
Perspectives, University of Lisbon Centre for English Studies, Lissabon, pp. 17–39.
Gunnarsson, B.-L. (2005), ‘The immigrant and the workplace: The societal frame’,
in B.-L. Gunnarsson (ed.), The Immigrant and the Workplace. (TeFa nr 41). Uppsala,
pp. 13–17.
Gunnarsson, B.-L. (2006), ‘Swedish companies and their multilingual practices’, in
J. C. Palmer-Silveira, M. F. Ruiz-Garrido and I. Fortanet-Gómez (eds), International
and Intercultural Business Communication. Theory, Research and Teaching. Bern: Peter
Lang, pp. 243–263.
Gunnarsson, B.-L. (2009a), Professional Discourse. (Continuum Discourse Series.)
London: Continuum.
Gunnarsson, B.-L. (2009b), ‘Multilingualism within transnational companies. An anal-
ysis of company policy and practice in a diversity perspective’, in H. Kelly-Holmes
and G. Mautner (eds), Language and the Market. Basingstoke and New York: Palgrave-
Macmillan.
Gunnarsson, B.-L. and Jämtelid, K. (1999), ‘Översättning och parallellskrivning –
flerspråkig textproduktion inom ett internationellt storföretag’, in Svenskans
beskrivning 23. Lund: Lund University Press, pp. 134–142.
Holmes, J. (2000), ‘Politeness, power and provocation: How humour functions in the
workplace’, Discourse Studies, 2, (2), 159–185.
Holmes, J. (2005), ‘Story-telling at work: A complex discursive resource for integrating
personal, professional and social identities’, Discourse Studies, 7, (6), pp. 671–700.
Holmes, J., Stubbe, M. and Vine, B. (1999), ‘Constructing professional identity: “Doing
power” in policy units’, in S. Sarangi and C. Roberts (eds), Talk, Work and Institu-
tional Order. Discourse in Medical, Mediation and Management Settings. Berlin and
New York: Mouton de Gruyter, pp. 351–385.
Holmes, J. and Stubbe, M. (2003), Power and Politeness in the Workplace. A Sociolinguistic
Study of Talk at Work. London: Longman.
Hull, G. (ed.) (1997), Changing Work, Changing Workers: Critical Perspective on Language
, Literacy and Skills. Albany: State University of New York Press.
Isaksson, M. (2005), ‘Ethos and pathos representations in mission statements: Identify-
ing virtues and emotions in an emerging business genre’, in A. Trosborg and
P. E. Flyvholm Jørgensen (eds), Business Discourse. Texts and Contexts, Bern, Berlin,
Frankfurt am Main: Peter Lang, pp. 111–138.
Johansson, C. (2003), Visioner och verkligheter. Kommunikationen om företagets strategier.
Department of Information Sciences, Uppsala University, Uppsala.
Jämtelid, K. (2001), ‘Multilingual text production at an international company’, in
F. Mayer (ed.), Language for Special Purposes: Perspectives for the New Millenium. Volume 2.
Tübingen: Gunter Narr Verlag, pp. 797–805.
Jämtelid, K. (2002), Texter och skrivande i en internationaliserad affärsvärld. Flerspråkig
textproduktion vid ett svenskt storföretag. Stockholm: Almqvist & Wiksell International.
Kankaanranta, A. (2005a), ‘English as a corporate language: Company-internal e-mail
messages written by Finns and Swedes’, in B.-L. Gunnarsson (ed.), Communication in
the Workplace. (TeFa nr 42), Uppsala, pp. 42–59.
Kankaanranta, A. (2005b), ‘Hej Seppo, could you pls comment on this!’ Internal Email Com-
munication in Lingua Franca English in a Multinational Company. Centre for Applied
Language Studies, University of Jyväskylä, Jyväskylä.
Karlsson, A. M. (2006), En arbetsdag i skriftsamhället. Ett etnografiskt perspektiv på skrifan-
vändning i vanliga yrken. (Småskrift utgiven av Språkrådet). Stockholm: Norstedts
akademiska förlag.
Karlsson, A.-M. (2007), ‘Text, situation, praktik. Om ramar och resurser för tolkning av
texter i arbetsrelaterade skrifthändelser’, in B.-L. Gunnarsson and A.-M. Karlsson
(eds), Ett vidgad textbegrepp. (TeFa nr 46). Uppsala, 27–40.
Kong, K. C. C. (2002), ‘Managing the ambiguous and conflicting identities of “upline”
and “downline” in a network marketing firm’, Discourse Studies, 4, (1), 49–74.
Landqvist, H. (2001), Råd och ruelse. Moral och samtalsstrategier i Giftinformations
centralens telefonrådgivning. (Skrifter utgivna av Institutionen för nordiska språk 55.)
Uppsala.
Lemke, J. L. (1999), ‘Discourse and organizational dynamics: Website communication
and institutional change’, Discourse & Society, 10 (1), 21–47.
LeVine, P. and Scollon, R. (ed.) (2004), Discourse and Technology: Multimodal Discourse
Analysis. Washington, DC: Georgetown University Press.
Linde, C. (1999), ‘The transformation of narrative syntax into institutional memory’,

Narrative Inquiry, 9, (1), 139–174.
Louhiala-Salminen, L. (1995), ‘Drop Me a Fax, Will You?’: A Study of Written Business
Communication. (Reports from the Department of English 10). Jyväskylä: University
of Jyvaskyla, 1–115.
Luzon, M. M. J. (2002), ‘A genre analysis of corporate home pages’, LSP and Profes-
sional Communication, 2 (1), 41–56.
Marra, M. and Holmes, J. (2004), ‘Affiliation workplace narratives and business reports:
Issues of definition’, Text, 24, (1), 59–78.
Nelson, M. (2007), ‘Professionella och socialiserande kommunikativa aktiviteter.
Andraspråkstalande kvinnor och män i arbete’, in B.-L. Gunnarsson, S. Entzenberger
and M. Ohlsson (eds), Språk och kön i nutida och historiskt perspektiv. (Skrifter utgivna
av Institutionen för nordiska språk 71.) Uppsala, 287–298.
Nelson, M. and Andersson, H. (2005), ‘Andraspråkstalare i arbetslivet. Data från
intervjuer och fallstudier’, in U. Börestam and B.-L. Gunnarsson (eds), Språk och
kultur i det multietniska Sverige (TeFa nr 44), Uppsala, pp. 84–97.
Nickerson, C. (2000), Playing the Corporate Language Game. An Investigation of the Genres
and Discourse Strategies in English Used by Dutch Writers Working in Multinational
Corporations. Amsterdam: Rodopi.
Norris, S. and Jones, R. H. (2005), Discourse in Action. Introducing mediated discourse
analysis. London and New York: Routledge.
Oliveira, M. C. L. (2004), ‘Language, technology and modernity: A study of interaction
in a Call Center’, in C. Gouveia, C. Silvestre, and L. Azuaga (eds), Discourse, Commu-
nication and the Enterprise. Linguistic Perspectives, University of Lisbon Centre for
English Studies, Lissabon, pp. 65–78.
Pan, Y., Scollon, S. W. and Scollon, R. (2002), Professional Communication in International
Settings. Malden, MA and Oxford: Blackwell Publishers.
Peters, T. J. and Waterman, R. H. (1982), In Search of Excellence: Lessons from America’s
Best-Run Companies. New York: Harper and Row.
Qvortrup, L. (2002), The Hypercomplex Society. New York: Peter Lang.
Sarangi, S. and Roberts, C. (eds) (1999), Talk, Work and Institutional Order. Discourse
in Medical, Mediation and Management Settings. Berlin and New York: Mouton de
Gruyter.
Silva, J. R. G., Oliveira, M. C. L. and Zaltzman, C. (2004), ‘Communication with
customers and the quality of public services. The case of a Brazilian company’,
in Gouveia, C., Silvestre, C. and L.Azuaga (eds), Discourse, Communication and the
Enterprise. Linguistic Perspectives, University of Lisbon Centre for English Studies,
Lissabon, pp. 43–63.
Swales. J. M. (1998), Other Floors, Other Voices: A Textography of a Small University Building.
Mahwah, NJ: Erlbaum.
Swales, J. M. and Rogers P. S. (1995), ‘Discourse and the Projection of Corporate
Culture: The Mission Statement’. Discourse & Society, 6, (2), 223–242.
Wiberg, M. (ed.) (2005), The Interaction Society. Practice, Theories and Supportive Technolo-
gies. Hershey, London, Melbourne, Singapore: Information Science Publishing.
CHAPTER
8
Political Discourse and Translation
Christina Schäffner
8.1 Introduction
A recent programme on BBC Radio 4 was devoted to the topic of pop music in
translation, reporting on translations of English pop songs of the 1960s, primarily
into German and French. The reporter commented, that, e.g., the Beatles’ song
‘She loves you’ had been recorded as a German version, but that a back-translation
from German into English revealed that the two texts were not exactly the same.
Despite such textual differences, the German version of the song was as huge a
success in Germany as the English version was in and beyond the United Kingdom.
This topic opens up two aspects for discussion: First, how aware are we that a text
we read or hear is a translation, i.e. how visible are translations? And secondly,
what do we expect translations to be like, i.e. what is the textual relationship
between an original text and its translation?
These questions can be asked with reference to practically all kinds of text
that are translated in various subject domains. In this chapter, they are explored
with reference to the domain of politics. A variety of texts are translated to fulfil
functions for political communication across languages, cultures and ideologies.
This chapter addresses the (in)visibility of translation in political communication,
textual profiles of translations (as compared to their source texts) and the role of
the socio-political contexts and conditions in which translations are produced.
The analyses are conducted primarily from the perspective of the discipline of
Translation Studies. At the end of the chapter, the role of Applied Linguistics
for Translation Studies is explored.
8.2 (In)Visibility of Translation in Political Communication
In late autumn 2007, the mass media reported that a Spanish court had found
21 people guilty of participating in the 2004 Madrid train bombings; seven other
defendants were acquitted. One of the chief suspects who was in jail in Italy had
allegedly bragged in a wiretapped phone conversation that the bombings had
been his idea. He was acquitted because his defence lawyer had argued successfully
that the transcripts of the tapes were mistranslated (see e.g. http://www.spiegel.
Political Discourse and Translation 143
de/international/europe/0,1518,514635,00.html, also http://www.npr.org/tem-

plates/story/story.php?storyId=15800071).
In autumn 2005, politicians and mass media around the world reacted very
critically to a speech by the Iranian president Mahmoud Ahmadinejad, given at
a conference in Tehran on 26 October 2005 and entitled ‘The World without
Zionism’. The British Prime Minister Tony Blair was reported to have ‘expressed
“revulsion” at the Iranian president’s assertion that he wanted Israel “wiped off the
map”’ (http://news.bbc.co.uk/1/hi/uk_politics/4380306.stm). This particular
phrase ‘wiped off the map’ was frequently commented on as being a mistransla-
tion. For example, the New York Times published an article that focused ‘exclusively
on the translation – or possible mistranslation – of the statement in order to deter-
mine if it constituted a threat against Israel and a call for war’ (reported on http://
disarmamentactivist.org/2006/06/, see also http://www.nytimes.com/2006/06/
11/weekinreview/11bronner.html?_r=1&oref=slogin).
In summer 2007, there was a debate in the UK mass media over rising costs
of public sector translation and interpreting. It was argued that the provision of
such services for the immigrant population was deterring community cohesion
in Britain. Ruth Kelly, the former Education Secretary, raised the question whether
translation had become a ‘crutch’ that discouraged integration (http://news.bbc.
co.uk/1/low/programmes/politics_show/6725673.stm).
What these examples illustrate is that translation seems to become visible in the
public eye if it is perceived as problematic. As readers, we are made aware of trans-
lation if some mistranslation had been discovered, leading to arguments about the
correct or incorrect meaning of a word, a sentence, or a text. Equally, thinking
of translation as a ‘crutch’ reflects an assumption of monolingualism being the
‘normal’ form of communication within a society.
In a positive sense, translation is immediately visible in the field of politics when
interpreting is involved. For example, during state visits, speeches by politicians
and/or press conferences are frequently interpreted, either simultaneously or
consecutively. In the case of press conferences, the interpreter often stands next to
the speaker and is thus visible as a person, in contrast to situations where only the
voice of the interpreter is heard over headphones, as normally happens in sessions
of the United Nations or the European Parliament.
There are, however, many more cases where translation is invisible. The general
public frequently encounters political discourse in translation without necessarily
noticing it. For example, newspapers regularly provide quotes of statements by
foreign politicians, without explicitly indicating that these politicians were actually
speaking in their own language, as can be seen in the following extract:
(1) ‘It is our common wish . . . that we get more transparency in financial mar-
kets,’ Merkel said after a regular meeting with Sarkozy at a government
guest house north of Berlin.
(http://www.iht.com/articles/ap/2007/09/10/europe/
EU-GEN-Germany-France.php)
144 Christina Schäffner
An American newspaper, the International Herald Tribune, quotes in English what

the German Chancellor had said, in all probability in German. The meeting she
had with the French President Sarkozy was interpreted, in accordance with com-
mon diplomatic practice, with Merkel speaking in German and Sarkozy in French.
In another example, the British weekly magazine The Economist in its regular com-
ment section ‘Charlemagne’ reports critically about comments by the Russian
President Putin on the relationships between Russia and the EU. We read:
(2) Meanwhile Russia is torn between wanting to be seen as, in Mr Putin’s

words, ‘a natural member of the European family’, and pride in being
an independent superpower, a fast-growing Eurasian giant and what
Dostoyevsky called ‘a sublime idea.’ (The Economist, 25 November 2006: 50)
There are two examples of quoting the original authors in direct speech,
although they did not speak in English in either case. The whole Economist column
is a critical engagement with a translation, without any explicit indication of this
fact. This practice reflects a general tendency of accepting translations as authen-
tic texts. Even if translations are visible and explicitly presented as translations, and
readers are aware of the fact that they are reading a translation, the general assump-
tion is that the translation is an exact replica of the original text. This deep-seated
expectation of faithfulness and fidelity to the original is also seen in a comment
by US President Bush, when at the end of an interpreted press conference at the
2007 G8-summit in Germany, he looked from the interpreter to the journalists and
asked: ‘Is this what I said?’ (as seen on BBC News reporting on 6 June 2007). It is
when we come across comments about mistranslation that we feel that this expec-
tation of what a translation will give us has been betrayed – which brings to mind
the age-old adage traduttore traditore, i.e. translators as betrayers and translations
as poor compromises, second best to the ‘holy’ original.
As said above, people engage with translations in a more or less direct way.
Politicians react to statements by other politicians as they were presented to them
in translation (e.g. Blair’s reaction to Ahmadinejad, quoted above). Media colum-
nists structure their own texts around statements by politicians presented in trans-
lation (e.g. example 2). Political scientists and lawyers debate the potential political
consequences of (the translation of) a statement (cf. the phrase ‘in order to
determine if it constituted a threat against Israel and a call for war’ quoted above).
Journalists report what foreign politicians said, which can be done by using direct
quotes (as in example 1), or even in an indirect way, as in the following example:
(3) Russia’s president, Vladimir Putin, said he would support Dmitry Medvedev,
first deputy prime minister, as the presidential candidate for his United
Russia party. That makes Mr Medvedev a near-certainty to take over. But
he promptly disappointed those who hope he will be more liberal than
his predecessor by saying he wanted Mr Putin as his prime minister, mean-
ing that Mr Putin is likely to call the shots. (The Economist, 15 December
2007: 9)
In this short news text, both Putin and Medvedev are quoted as having said
something, without any indication of when and where they said this and in which
language.
The question of visibility or invisibility of translation needs to be linked to the
contexts in which translations are made available and the functions they fulfil for
the respective communicative purpose. Particularly in the case of mass media, only
translated extracts of original texts, such as speeches, are included in journalistic
articles. Journalistic texts too, especially editorials and comments, can have an
impact on public opinion about politics and also on policy making. This raises the
question of how political communication, political discourse and political texts
can be defined.
As argued elsewhere (Chilton and Schäffner, 1997: 206), ‘politics cannot be
conducted without language, and it is probably the case that the use of language
in the constitution of social groups leads to what we call “politics” in a broad sense’.
Subsequently, all ‘those actions (linguistic or other) which involve power, or its
inverse, resistance’ can be defined as potentially political (Chilton and Schäffner,
1997: 212). Political discourse as a complex form of human activity is realized in a
variety of discourse types (or genres), whose discourse organization and textual
structure is determined by the respective political situations and processes (discur-
sive practices). Burkhardt (1996) suggests a broad distinction between communi-
cating about politics, political discourse in mass media and political communication,
and Lilleker (2006: 1) highlights that non-state actors (such as pressure groups)
need to be added to the main actors in political communication, i.e. the state, the
media and the public. Although it is useful to maintain a distinction between insti-
tutional politics and everyday politics, this chapter only addresses institutionalized
forms of political discourse, i.e. texts that originate in political or media institu-
tions. Here we can broadly differentiate between texts that are instrumental in
policy making and thus produced by and addressed to politicians, and texts that
communicate, explain and justify political decisions, produced by politicians,
political scientists or journalists and addressed to the general public. In both cases,
communication may cross linguistic and cultural boundaries, which means that
translation is involved.
Translation is thus relevant not only for international and foreign affairs, but also
for domestic affairs. In respect of international politics, translation plays a significant
role in international organizations such as the United Nations or the European
Union institutions. Concerning the foreign policy of individual states, translation
becomes relevant, e.g., for signing bilateral treaties and for delivering speeches
during state visits. Translations of such speeches are made available on websites of
governments or of embassies. In this way, the government communicates its politi-
cal aims and decisions to the outside world. But political aims and decisions of for-
eign countries are also presented to home governments in translation. For example,
the BBC Monitoring Service translates texts into English for the UK government.
Thus, translation plays a role for both exporting and importing political texts.
As far as domestic politics is concerned, translation is relevant in multilingual
societies for communicating with a multilingual and multiethnic public. Although the
United Kingdom is not officially a multilingual country (cf. Blackledge, 2005),

a variety of documents are translated into community languages. For example, a
leaflet prepared by the UK Government and entitled ‘Preparing for emergencies.
What you need to know’ is available, in addition to English and Welsh, in Bengali,
Chinese, Arabic, Somali, French, Urdu, Punjabi, Gujarati, and the text can also be
accessed via a website in Greek, Kurdish, Farsi, Turkish, Vietnamese and Hindi.
This leaflet includes a section on helping to prevent a terrorist attack, which means
that this leaflet too can be regarded as an example of political discourse in the
widest sense.
In sum: translation is an integral part of politics. Which texts get translated and
from and into which languages is itself already a political decision. Translation is
not simply a linguistic activity, but rather a social-political practice. Translators
work in social-political and historical contexts, their activity is embedded in and
determined by social, institutional, ideological norms, conditions and constraints.
Translational behaviour is thus contextualized social behaviour, with the act of
translation, i.e. the cognitive aspects of translation as a decision-making process,
embedded in a translation event, i.e. the social, historical, cultural, ideological
context (cf. Toury, 1995). By linking translations (as products) to their social con-
texts, causes and effects of translations can be discovered (cf. Chesterman, 1998).
In the following sections, I illustrate the link between translation profiles and
the social, institutional, ideological conditions of text production with reference
to texts which can be described as typical (or: prototypical) examples of political
discourse: the Draft Treaty establishing a Constitution for Europe, political speeches
and press conferences. Translation for global news, which has only recently received
more attention within Translation Studies will not be considered here (but see
Bassnett, 2004; Bielsa, 2007; Holland, 2006; Kang, 2007; Schäffner, 2005; on previ-
ous work on the topic of translation and politics see Schäffner, 2007).
8.3 Translation Profiles and Conditions of Text Production
8.3.1 Translation in and for Multilingual Institutions
International multilingual organizations, such as the United Nations and the

European Union, have their own language and translation policies, and also their
own translation departments. In the United Nations Organization, the working
languages are English, French, Spanish, Russian, Chinese, in contrast to the UN
official languages English and French. For the European Union, all national lan-
guages of the member states are official languages due to the EU’s language policy
(see, e.g., Tosi, 2002, Wagner et al., 2002). Every citizen of the EU has the right
to use their own national language in communicating with the EU institutions.
Due to this policy, translation and interpreting services have expanded massively.
Koskinen (2000) divided translations in the EU institutions into two groups: intra-
cultural communication and intercultural communication, depending on the pro-
ducers, addressees and functions of the texts (e.g. intrainstitutional translation
intended for internal use within the same institution, or translations produced
within one of the EU institutions and intended for communication with the general
public). For a significant number of official texts, the principle of equal authentic-
ity of all languages and texts applies. This policy has led to the phenomenon that
translation, although a huge enterprise, is invisible in the official discourse (e.g. the
Council Regulation No 1 speaks of ‘language versions’, but not explicitly of ‘trans-
lation’). The political aim of equal authenticity results in specific linguistic features
of the various language versions, as illustrated by Seymour (2002) and Koskinen
(2000, who speaks of an ‘illusion of identity’). Once published, it is impossible to
identify one text as an original source text. However, as Koskinen (2001) explains,
the drafting process is much more complex, and it is therefore more accurate to
speak of a textual network comprising all different language versions of the text
that may have functioned as source texts during several stages of translation.
The close structural, lexical and grammatical similarity between the various lan-
guage versions can be illustrated with an extract from the English, German, French
and Spanish version of the Preamble of Part II of the Draft Constitutional Treaty
(The Charter of Fundamental Rights of the Union). Each sentence in the English
version corresponds to exactly one sentence in the French and Spanish versions,
with only the German version having split the second sentence into two sentences.
The syntactic structure of the sentences is identical in all four versions, but with
the position of the main clause and the sub-clause reversed in the last sentence of
the Spanish version.
(3a) Part II
The Charter of Fundamental Rights of the Union
Preamble
The peoples of Europe, in creating an ever closer union among them, are
resolved to share a peaceful future based on common values.
Conscious of its spiritual and moral heritage, the Union is founded on
the indivisible, universal values of human dignity, freedom, equality and
solidarity; it is based on the principles of democracy and the rule of law.
It places the individual at the heart of its activities, by establishing the
citizenship of the Union and by creating an area of freedom, security and
justice.
(http://eur-lex.europa.eu/JOHtml.do?uri=OJ:C:2004:310:
SOM:EN:HTML)
(3b) Teil II
Die Charta der Grundrechte der Union
Präambel
Die Völker Europas sind entschlossen, auf der Grundlage gemeinsamer
Werte eine friedliche Zukunft zu teilen, indem sie sich zu einer immer
engeren Union verbinden.
In dem Bewusstsein ihres geistig-religiösen und sittlichen Erbes gründet
sich die Union auf die unteilbaren und universellen Werte der Würde des
Menschen, der Freiheit, der Gleichheit und der Solidarität. Sie beruht auf
den Grundsätzen der Demokratie und der Rechtsstaatlichkeit. Sie stellt
den Menschen in den Mittelpunkt ihres Handelns, indem sie die Unions-
bürgerschaft und einen Raum der Freiheit, der Sicherheit und des Rechts
begründet.
SOM:DE:HTML)
(3c) Partie II
La Charte des Droits Fundamentaux de l’Union
Préambule
Les peuples d’Europe, en établissant entre eux une union sans cesse plus
étroite, ont décidé de partager un avenir pacifique fondé sur des valeurs
communes.
Consciente de son patrimoine spirituel et moral, l’Union se fonde sur
les valeurs indivisibles et universelles de dignité humaine, de liberté,
d’égalité et de solidarité; elle repose sur le principe de la démocratie et le
principe de l’État de droit. Elle place la personne au cœur de son action
en instituant la citoyenneté de l’Union et en créant un espace de liberté,
de sécurité et de justice.
SOM:FR:HTML)
(3d) Parte II
Carta de los Derechos Fundamentales de la Unión
Preámbulo
Los pueblos de Europa, al crear entre sí una unión cada vez más estrecha,
han decidido compartir un porvenir pacífico basado en valores comunes.
Consciente de su patrimonio espiritual y moral, la Unión está fundada
sobre los valores indivisibles y universales de la dignidad humana, la liber-
tad, la igualdad y la solidaridad, y se basa en los principios de la demo-
cracia y el Estado de Derecho. Al instituir la ciudadanía de la Unión y
crear un espacio de libertad, seguridad y justicia, sitúa a la persona en el
centro de su actuación.
SOM:ES:HTML)
As said above, it is the language policy of the EU which requires a close similarity
of the different versions. The difficulty, or rather, the impossibility, to identify
which language version might have been based on which other one can be illus-
trated with reference to the use of the heart metaphor in the English and French
versions (‘at the heart of its activities’, ‘au cœur’), whereas the German and
Spanish versions have a more general reference to the centre (‘in den Mittelpunkt’,
‘en el centro’). The German version is more specific with ‘geistig-religiösen und
sittlichen Erbes’ (literally: spiritual-religious and moral-ethical heritage) compared
with the other three versions (‘spiritual and moral heritage’, ‘patrimoine spirituel
et moral’, ‘patrimonio espiritual y moral’). This difference may be explained with
reference to differences in the linguistic systems, with ‘geistig’ less directly convey-
ing the sense of religion than ‘spiritual’, ‘spirituel’ and ‘espiritual’ do. However,
there had been a long and controversial debate among the member states on
whether the Preamble should include a reference to the role of religion in Europe’s
history or not. The decision to add ‘religiös’ in the German version could thus also
be interpreted as a conscious translation decision, signalling a political compromise.
In the context of European Union politics, it happens as well that several mem-
ber states join forces and present a common position to the other member states.
This was the case in 2003, when the heads of government of Germany, France
and the United Kingdom addressed the President of the European Commission in
a joint letter (http://66.102.11.104/search?q=cache:kvXx7GGoLKgJ:www.amba-
france-de.org/aktuelle.php3%3Factu%3D1027+Brief+%2B+Chirac+%2B+Blair+
%2B+prodi&hl=en&ie=UTF-8; http://www.ambafrance-uk.org/Risks-of-de-indus-
trialization.html). This letter was made available in German, English and French
and published on the websites of the three governments in their own languages,
respectively. Language- and culture-specific conventions can be seen in the forms
of address (‘Sehr geehrter Herr Präsident, Dear President, Monsieur le Président’)
and the closing formula (‘Mit freundlichen Grüßen, Yours sincerely, Nous vous
prions d’agréer, Monsieur le Président, l’expression de notre haute considéra-
tion’). In the introduction to the letter, the sequence in listing the names is differ-
ent (for the German version: Jacques Chirac, Gerhard Schröder, Tony Blair; for
the English and French versions: Jacques Chirac, Tony Blair, Gerhard Schröder,
with the alternative spelling Schroeder). When the names are repeated after the
closing formula, the name of the head of government of the respective country
comes first (the French version did not repeat the names at the end).
(4a) In einem gemeinsamen Brief wandten sich Staatspräsident Jacques

Chirac, Bundeskanzler Gerhard Schröder und Premierminister Tony Blair
bei ihrem Treffen am 20. September 2003 in Berlin an den Präsidenten
der Europäischen Kommission, Romano Prodi:
Sehr geehrter Herr Präsident,
der Europäische Rat hat auf unsere Initiative am 21. März 2003 ein
klares Signal zur Stärkung der industriellen Wettbewerbsfähigkeit der EU
gegeben. Wir waren uns einig, dass wir den Verwaltungsaufwand unserer
Wirtschaft verringern und den Regelungsrahmen, in dem sich unsere
einem harten Wettbewerb ausgesetzten Unternehmen bewegen müssen,
entscheidend verbessern müssen. [. . .]
Mit freundlichen Grüßen
Gerhard Schröder Jacques Chirac Tony Blair
(4b) Risks of de-industrialization – Letter from M. Jacques Chirac, President of

the Republic, Mr Tony Blair, Prime Minister of the United Kingdom of Great
Britain and Northern Ireland, and Mr Gerhard Schröder, Chancellor of
the Federal Republic of Germany, to Mr Romano Prodi, President of the

European Commission.
Berlin, 20 September 2003
Dear President,
On our initiative on 21 March 2003, the European Council sent out a
clear signal for strengthening the industrial competitiveness of the EU. We
agreed that we had to reduce the bureaucracy that European companies
encounter and decisively improve the regulatory framework within which
our companies, faced with strong competition, must manoeuvre. [. . .]
Yours sincerely.
Tony Blair, Jacques Chirac, Gerhard Schröder
(4c) Risques de Desindustrialisation

Lettre du Président de la République, M. Jacques Chirac, du Premier
Ministre du Royaume-Uni, M. Tony Blair, du Chancelier de la République
Fédérale d’Allemagne, M. Gerhard Schroeder, au Président de la Commis-
son Européenne, M. Romano Prodi
(Berlin, 20 septembre 2003)
Monsieur le Président,
A notre initiative, le Conseil européen a, le 21 mars 2003, émis un signal
clair en faveur du renforcement de la compétitivité industrielle de l’Union.
Nous avons reconnu la nécessité de réduire la charge bureaucratique qui
pèse sur les entreprises européennes et d’améliorer de manière décisive le
cadre réglementaire dans lequel celles-ci, confrontées à une vive concur-
rence, doivent opérer. [. . .]
Nous vous prions d’agréer, Monsieur le Président, l’expression de notre
haute considération.
Here too, it is impossible to say with absolute certainty whether only one text
served as a source text. However, a comparison of the three versions shows rela-
tively close similarities between the English and French versions (e.g. ‘sent out
a clear signal’ and ‘émis un signal clair’; both starting in exactly the same way
with the prepositional phrase ‘on our initiative’/‘à notre initiative’, ‘European
companies’/‘les entreprises européennes’ and a very similar syntactic structure)
in contrast to the German version.
It could thus be speculated that the complexity of the drafting processes, which
Koskinen emphasized with respect to multilingual versions of EU texts of a more
binding nature, which are agreed only after lengthy negotiations, does not apply
in the same way to multilingual texts which are meant to fulfil a specific and short-
term purpose and do not have a legally binding status. However, other politicians
in other countries may be interested in being informed of this letter as well, as may
journalists who wish to write about this initiative in the press. For producing
versions of this letter then in languages other than German, English or French, it
is highly unlikely that all three texts will be used to function equally as source texts
for another translation.
8.3.2 Political Speeches and Statements
Politicians give speeches on various occasions, both in their home country, on state
visits to another country or at sessions of international organizations, such as in the
United Nations or the European Parliament. The text of such speeches is often
made available to journalists in advance. In that case, the text is usually preceded
by the statement ‘Check against delivery’, indicating that journalists must only
report on the speech as it was actually delivered. Translations of speeches too are
sometimes made available in advance, and this is usually indicated by the phrase
‘Advance translation’.
Increasingly, such speeches are made available on websites as well, in the origi-
nal version and sometimes also in translation. The decision to have a multilingual
website of a government, the choice of languages and the selection of speeches for
translation are all politically motivated. Not all speeches get translated, and embas-
sies often publish more speeches, both in the original and in translation, than gov-
ernments do on their own homepages.
This raises the question of the addressees of such speeches. The concept of
audience design can be applied here. Bell (1984) and Mason (2000) distinguish
between addressees (those whose presence is known, who are ratified participants
in an exchange and who are directly addressed), auditors (those who are known,
ratified, but not directly addressed), overhearers (known but not ratified parti-
cipants and not addressed), and eavesdroppers (those whose presence is not
even known). The two examples below are a New Year’s message and a statement
in Parliament. The first case is Tony Blair’s New Year’s message, delivered on
31 December 2003. The initial addressees are people living in the United Kingdom.
The British Embassy in Germany had made this address available on its website in
both English and German. The addressees of the English text can be identified as
UK citizens living in Germany. When the address was initially delivered, German
speakers were not thought of at all. They could at best be seen as eavesdroppers.
It is as a result of the decision to produce a German translation of this text that a
new group of addressees came into being. However, the function of the text has
changed: for the addressees of the translation, the text is fulfilling an informative
function, i.e. informing readers in another culture of what the British Prime
Minister said to his own citizens, and how he wanted to persuade them to accept
his arguments. Differences in the background knowledge of German addressees
concerning issues specific to the UK culture were taken into account for produc-
ing the translation, as can be seen in translation strategies of explanation, explici-
tation, information addition. For example, in a section on economic growth, ‘Bank
of England’, ‘windfall tax’ and ‘New Deal’ have been explained and/or specified
in the German translation, with inverted commas signalling the English words, cf.:
(5) None of this has happened by accident but because of the hard choices that
made it possible – Bank of England independence, tough fiscal rules, two
years of tight spending, the windfall tax on the privatized utilities paying for
the New Deal.
All das verdanken wir nicht dem Zufall, sondern harten Entscheidungen:
Unabhängigkeit unserer Zentralbank – der Bank von England – (literally:
our Central Bank – the Bank of England –), harte steuerpolitische Maßnah-
men, zwei Jahre strenger Ausgabendisziplin, die ‘windfall’ Steuer, mit der
unerwartete Gewinne aus privatisierten Unternehmen der öffentlichen
Versorgung abgeschöpft und für die Politik des ‘New Deal’ eingesetzt wur-
den (literally: the windfall tax with which unexpected profit from privatized
utilities in the public services was taken and made available for the policy of
the New Deal).
(http://www.britischebotschaft.de/en/news/items/040101.htm,
accessed 30 January 2004)
The speeches we find on government websites and embassy websites are nor-
mally translated with attention to a fluent style in the target text. In the case of this
particular New Year message, the German version has added two sentences at the
very beginning, thus contextualizing the information. This strategy resulted in
enhanced coherence.
(5a) The decision to go to war in Iraq was the most difficult of all.
In den letzten Jahren haben wir einige sehr harte Entscheidungen
getroffen, besonders in den letzten zwölf Monaten. Sie waren notwendig,
und ich glaube, sie zeigen erste Erfolge.
Die schwierigste von allen war die Entscheidung, Krieg gegen den Irak
zu führen.
(Literally: Within the last years we have taken some difficult decisions,
particularly in the last twelve months. They were necessary, and I think
they show the first successes.
The most difficult of all was the decision to go to war in Iraq.)
It is peculiar even for the English original to start without a form of address.
However, the lack of a formal address can be explained with reference to the
recontextualization of the speech. That is, the message was made available on the
Embassy website shortly after its delivery, i.e. the immediacy of the Prime Minister
addressing the public is no longer required. It is a rather frequently occurring
feature that forms of address do not show up when speeches are made available on
the internet subsequent to the original context of delivery. However, it is interest-
ing to see that in some cases forms of address have been added to the translated
versions, or have been expanded, as in the excerpts from two other speeches below:
(6) Good morning. – Guten Morgen, meine sehr geehrten Damen und Herren!
(literally: Good morning ladies and gentlemen)
(http://www.britischebotschaft.de/en/news/items/071210.htm)
Let me start by thanking you all for coming and for this invitation to speak
here today. – Meine Damen und Herren! Ich möchte zunächst Ihnen allen
für Ihr Kommen und für die Einladung danken, heute hier zu sprechen.
(‘Ladies and gentlemen’ was added)
(http://www.britischebotschaft.de/de/news/
items/070523a.htm)
The production of fluent target texts seems to be the normal institutional prac-
tice as well when it comes to translating speeches by politicians, which were deliv-
ered abroad during state visits, or in the European Parliament. Although such
speeches are usually interpreted simultaneously, advance translations are often
prepared and distributed. The addressees of such a speech are MPs and politicians
present in the Parliament. Other people who are present, including journalists,
are auditors, whereas readers of the texts as made available on the internet may be
seen as overhearers or eavesdroppers. Addressees of the translated versions are the
politicians who are present, which explains the fluency of the target texts. This can
be illustrated with the first paragraph of the original German version of a speech
delivered by the German Chancellor Angela Merkel to the European Parliament
on 13 February 2007 and its English and French translations:
(7a) Herr Präsident, Sie gehören zu den Parlamentariern der ersten Stunde im
Europäischen Parlament. Sie haben es eben gesagt, Sie haben Erfahrung
seit 1979. Ich glaube, man kann sagen, dass Sie einen bemerkenswerten
Aufstieg eines Parlamentes, das in Kinderschuhen begonnen hat, erlebt
und mitgestaltet haben. Es hat sich zu einem sehr emanzipierten
Europäischen Parlament entwickelt – mit selbstbewussten Abgeordneten,
klaren Parteistrukturen und Fraktionsbildungen. Damit wurde es ein
anspruchsvoller, aber nicht mehr wegzudenkender Partner im europäischen
Diskurs.
(http://www.bundesregierung.de/nn_1498/Content/DE/Rede/
2007/02/2007-02-13-rede-merkel-ep.html)
(7b) Mr President, you are one of the parliamentarians who witnessed the first
hours of the European Parliament. You have just said that your experience
goes back to 1979. I think we could say that you have experienced and
helped influence the remarkable rise of a Parliament from its infancy. It
has evolved into a very emancipated European Parliament – with confi-
dent deputies, clear party structures and parliamentary groups. This has
made it a critical but now irreplaceable partner in European debate.
(http://www.bundesregierung.de/nn_6566/Content/EN/Reden/
2007/02/2007-02-13-rede-merkel-ep__en.html)
(7c) Monsieur le Président, vous faites partie des parlementaires de la première

heure au Parlement européen. Comme vous venez de le dire, vous avez
acquis de l’expérience depuis 1979. Il me semble pouvoir dire que vous
avez vécu et contribué à forger la remarquable ascension d’un parlement
depuis ses premiers balbutiements. Ce parlement est devenu un Parlement
européen très émancipé, avec des députés sûrs d’eux, des structures de parti
et des groupes parlementaires bien définis. C’est aujourd’hui un partenaire
exigeant, mais qui ne pourrait plus être absent du débat européen.
(http://www.bundesregierung.de/nn_5846/Content/FR/Reden/
2007/02/2007-02-13-rede-merkle-ep-fr.html)
These findings are in line with those of Calzada Pérez (2001) who analysed
translated speeches in the European Parliament (Spanish – English). Her analysis,
carried out through surface description, illocutionary explanation and (socio-
political) perlocutionary explanation, revealed a broad variety of translational
shifts which were intended to help target texts to be more readable, thus contra-
dicting Koskinen’s (2000) findings of target texts mirroring source texts in their
linguistic structure. With translation not consisting simply in replacing lexical and
grammatical units in the source text by equivalent units in the target text, it is quite
normal that target texts reveal differences in syntactic structures and lexical
choices. One quite common syntactic translation strategy is the shift from a clause
in passive voice to an active one, or vice versa (what Chesterman, 1997 calls clause
structure change). Such a syntactic change can be seen in the following example,
the first paragraphs of a joint article by the French Foreign Minister Bernard
Kouchner and the British Foreign Secretary David Miliband which was published
in the International Herald Tribune, 14 October 2007, and subsequently made avai-
lable in English and in German translation on the website of the UK Embassy in
Germany, cf.:
(8a) The world has reacted with horror to the Burmese regime’s brutal crack-
down against its own people. Monks, nuns and ordinary citizens took to
the streets peacefully in protest at the deterioration of the economic situa-
tion in the country. They were met with guns and batons.
We cannot know for sure the number of those who were killed, but it is
likely to be many more than the regime is willing to admit. The where-
abouts and welfare of many who have been detained remain uncertain.
Meanwhile, the persecution continues: The security forces carry out new
raids and new arrests every night.
(http://www.iht.com/articles/2007/10/14/opinion/edkouchner.php,
also: http://www.britischebotschaft.de/en/news/items/071014.htm)
(8b) Mit Entsetzen hat die Welt auf das brutale Vorgehen des Regimes in Birma
gegen sein eigenes Volk reagiert. Mönche, Nonnen und normale Bürger-
innen und Bürger waren auf die Straße gegangen, um friedlich gegen die
Verschlechterung ihrer wirtschaftlichen Lage zu protestieren. Das Regime
reagierte mit Gewehren und Schlagstöcken.
Die genaue Zahl der Todesopfer kennen wir nicht, aber sie liegt wahr-
scheinlich viel höher, als das Regime zuzugeben bereit ist. Der Aufenthalts-
ort vieler Verhafteten und deren weiteres Schicksal sind nach wie vor
unbekannt. Inzwischen setzt das Regime seine Repressalien fort: Die
Sicherheitskräfte führen jede Nacht neue Razzien und Verhaftungen

durch.
(http://www.britischebotschaft.de/de/news/items/071014.htm)
Whereas the agent is left implicit in ‘[t]hey were met with guns and batons’ and
in ‘the persecution continues’, the German version has an explicit reference to ‘the
regime’ in both cases (‘Das Regime reagierte mit Gewehren und Schlagstöcken –
The regime reacted with guns and batons; setzt das Regime seine Repressalien
fort – The regime is continuing its repressive measures’). It could be argued that
as a result of such a strategy, the active role of the regime has been put more in
the foreground. The French version of this text, available on the website of the
French embassy in the United Kingdom, has the same passive structure as the
English text, cf.:
(8c) La brutalité avec laquelle le régime birman a réprimé son propre peuple
nous a tous horrifiés. Des moines, des nonnes et des citoyens ordinaires,
descendus pacifiquement dans la rue pour protester contre la situation
économique du pays, ont été accueillis à coups de fusils et de gourdins.
Le chiffre exact des morts et des blessés reste inconnu et il est probable-
ment beaucoup plus élevé que celui avancé par les autorités. L’état de
santé des nombreux prisonniers suscite l’inquiétude. Et, pendant ce temps,
les persécutions continuent: les forces de sécurité procèdent chaque nuit
à de nouveaux raids et à de nouvelles arrestations.
(http://www.ambafrance-uk.org/Renforcer-la-pression-sur-la.html)
Since the complete text does not reflect a significantly foregrounded role of the
regime, it would be inappropriate to say that the German translator chose such a
strategy deliberately in order to show his or her own attitude and engagement.
One area where certain translation strategies are chosen deliberately in order to
promote a specific ideological agenda is feminism, with translators as intervention-
ists, e.g. ‘feminizing’ words in the target text (e.g. ‘auteure’, the French word for
a female author becomes ‘auther’, rather than ‘author’, and ‘amante’, a female
lover, becomes ‘shelove’, Simon, 1996: 21). Engagement on the part of translators
is a topic which has recently seen more attention in Translation Studies (e.g. Baker,
2006). Ideological aspects of translation are also clearly highlighted by Tymoczko
and Gentzler in the following quote:
Translation thus is not simply an act of faithful reproduction but, rather, a delib-
erate and conscious act of selection, assemblage, structuration and fabrication –
and even, in some cases of falsification, refusal of information, counterfeiting,
and the creation of secret codes. (Tymoczko and Gentzler, 2002: xxi)
On the basis of the final example, press conferences, I will show how institu-
tional practices result in a particular way of presenting (and constructing)
politicians.
8.3.3 Addressing the Public: Press Conferences
It is a usual procedure that heads of state or government address representatives

of the media after state visits. At such press conferences, summarizing comments
are followed by a question and answer session. The texts of such press conferences
are often made available on websites as well. Either simultaneous or consecutive
interpreting is practised at such press conferences. Different governments have
different practices in signalling (or hiding, as the case may be) the involvement of
translation or interpreting. For example, the UK government website makes press
conferences with foreign visitors available almost exclusively in English, there is
rarely an indication of the fact that visitors did not speak in English and that inter-
preting and/or translation was involved. For example:
(9) Press Conference of the UK Prime Minister Gordon Brown and the
German Chancellor Angela Merkel, 22 August 2007, London:
Prime Minister: Can I say what a pleasure it has been to have Chancellor
Merkel here in Downing Street. [. . .]
Chancellor Merkel: Thank you very much. It is with great pleasure that
I have come here today to London. [. . .]
(http://www.number10.gov.uk/output/Page12905.asp)
The same applies if a press conference was held outside the United Kingdom,
e.g. a press conference by Prime Minister Tony Blair, German Chancellor Gerhard
Schroeder and French President Jacques Chirac, 20 September 2003, in Berlin
(the meeting which resulted in the joint letter to Romano Prodi, see example 4
above).
(10) Chancellor Schroeder:

Good Afternoon Ladies and Gentlemen. I would like to give you a few
pieces of information on the results of this informal meeting. [. . .]
President Chirac:
Let me first of all thank once again the Chancellor for his very warm
welcome, very warm and very pleasant welcome. As the Chancellor rightly
pointed out, [. . .]
Prime Minister:
For my part, first of all let me give my thanks to Chancellor Schroeder for
hosting this meeting and say what a constructive meeting I found it. [. . .]
Even when there is an explicit reference to the use of another language, as in

the first turn of the Portuguese Prime Minister at a press conference on 9 July 2007
in London, there is no explicit indication of translation or interpreting:
(11) Prime Minister:

Ladies and Gentlemen I am delighted that my first foreign visitor to No 10
should be the President of the European Union Council of Ministers and
the Prime Minister of Portugal. [. . .]
Mr Socrates:
Thank you Prime Minister. I will speak in Portuguese, if you don’t mind.
It will be better for me and better for you.
I would like to start by thanking you Prime Minister for inviting me here.
[. . .]
The website of the German government presents press conferences exclusively

in German, but it is in most (at least more recent) cases indicated that the tran-
scripts of the statements in a language other than German are based on simultane-
ous (or occasionally consecutive) interpreting. As the analysis of a number of
press conferences has shown, some revision process is in place for the German
government’s website. The transcripts reflect that the orally delivered contribu-
tions by the speakers and the interpreters are grammatically and stylistically
improved to a certain extent for the written texts. This can be seen in an extract
from a press conference with the German Chancellor Angela Merkel and George
Bush in Stralsund, Germany, on 13 July 2006. The transcript on the website of the
White House reflects the typical features of oral speech (e.g. false starts, self correc-
tions), whereas in German, Bush’s discourse is fluent and grammatically correct.
(12) Bush:
You know, on the Iranian issue, for example, the last time that we were
together we talked – spent a lot of time on Iran, and the Chancellor was
wondering whether or not the United States would ever come to the table
to negotiate with the Iranians. You made that pretty clear to me that you
thought it was something – an option we ought to consider, which I did.
And I made it clear to the Iranians that if they were to do what they said
they would do, which is to stop enrichment in a verifiable fashion, we’re
more than pleased to come back to the table. [. . .]
(http://www.whitehouse.gov/news/releases/2006/07/20060713-4.html)
Das tun wir auch, was den Iran betrifft. Bei unserem letzten Treffen haben
wir mit dieser Frage sehr viel Zeit verbracht. Dabei wurde die Frage gestellt:
Werden sich die Vereinigten Staaten jemals zum Verhandlungstisch bege-
ben? Die Bundeskanzlerin hat mich auch dazu aufgefordert, darüber
nachzudenken. Ich habe dann Folgendes gesagt: Wenn die Iraner nach-
weislich mit der Urananreicherung aufhören, dann werden wir zum
Verhandlungstisch zurückkehren. (Literally: We do the same in respect
of Iran. At our last meeting we devoted much time to this issue. Then the
question was asked: Will the United States ever come to the table to nego-
tiate? The Chancellor asked me to reflect about this. I then said the follow-
ing: If there is evidence that the Iranians stop uranium enrichment, then
we will return to the table to negotiate.)
(http://www.bundesregierung.de/nn_1516/Content/
DE/Mitschrift/Pressekonferenzen/2006/07/
2006-07-13pressekonferenz-merkel-bush.html)
Academics or students who want to do an analysis of press conferences will inevi-

tably arrive at different evaluations of the politicians’ performance. In November
2007, the German Chancellor Merkel visited US President Bush. The press confer-
ence of 10 November 2007 is available in full on the website of the White House,
but only the German text by Merkel is available on the website of the German
Government. There is a note at the beginning of the text, saying that the English
parts of the press conference were not translated. The whole transcript then starts
by indicating Bush as the first speaker, with information in brackets after his name
saying that he was speaking in English. Then we have a long statement by Merkel
in German (in the transcript, ‘P’ stands for ‘Präsident’ and ‘BK’in’ for ‘Bundeskanz-
lerin’, the female form of Federal Chancellor):
(13) P Bush: (auf Englisch)

BK’in Merkel: Sehr geehrter Herr Präsident, lieber George, ich möchte
mich zuerst für diese Möglichkeit bedanken, [. . .]
Exchanges between Bush and the journalists are again just given by saying
Question (in English), Bush (in English). But since some questions are addressed
to both Bush and Merkel, answers by Merkel are incoherent for a German reader
who does not know what the actual question was. The very end of the press confer-
ence too must leave the German reader wondering what was going on. The German
website provides the following information:
P Bush: (auf Englisch)

BK’in Merkel: Das ist für mich als eine Hamburgerin natürlich eine wunderbare
Sache.
(Literally: This is something wonderful for me as a woman from Hamburg.)
The White House website has a complete version of the press conference in
English. All of Merkel’s statements and answers are introduced by the information
‘as translated’ in brackets. The closing part of the conference reads as follows on
the White House website:
Chancellor Merkel: (As translated.) Well, from my side, [. . .]

President Bush: I’m now going to go feed the Chancellor a hamburger. (Laughter.)
Right here, Crawford, Texas. No, well, I mean back over there. Thank you all.
Chancellor Merkel: Obviously, for me, as a person who originally came from
Hamburg –
President Bush: Yes.
Chancellor Merkel: – it’s even more important.
President Bush: Hamburger.
Thank you.
Chancellor Merkel: Thank you.
Comparing the English version of Merkel’s comments as published on the White

House website to the German text on the website of the German government, we
notice again some differences, especially in style. This is illustrated in the following
extract:
BK’in Merkel: Sehr geehrter Herr Präsident, lieber George, ich möchte mich
zuerst für diese Möglichkeit bedanken, hier in Texas diese Gespräche zu füh-
ren, auch im Namen meines Mannes. Es ist ein wunderschönes Fleckchen Erde,
wie man in Deutschland sagen würde, das etwas von der amerikanischen Weite
und der Vielfalt der Landschaft zeigt. Herzlichen Dank dafür, dass wir heute
Morgen die Möglichkeit zum Spaziergang hatten. Das war eine wirklich neuar-
tige Erfahrung.
(Literally: Dear Mr President, dear George, first of all, allow me to thank you
for the possibility to have these talks here in Texas, also on behalf of my hus-
band. This is a very beautiful spot on earth, as we would say in Germany, which
shows us something of the American vastness and variety of the countryside.
Thank you very much for the opportunity we had this morning to go for a walk.
This was a totally new experience.)
Chancellor Merkel: (As translated.) Well, yes, thank you very much, Mr. President,
dear George. First of all, allow me to thank you very warmly for the possibility to
meet with you here in Texas and to have this exchange of views. I would also like
to extend this word of gratitude to you on behalf of my husband, who accompa-
nies me here to this, what we also in Germany would call a very beautiful spot, a
very beautiful part of this planet, of this world. It enables us to appreciate a little
bit the vastness of the territory here, and also the beauty and the sheer variety of
species that you have here.
So we again were able to see this for ourselves this morning. Thank you again
for making this possible to have this stroll with you and to appreciate the beauty
of this part, and to have again an exchange of views on a number of subjects.
I said above that written versions of oral statements have been stylistically
enhanced for publication on the website of the German government. The practice
for the White House is different. The differences in style which we see above can
be attributed to the fact that Merkel’s turns were interpreted consecutively, as
can be seen from the video of the press conference which can be accessed via a
link from the White House website. Hesitations, repetitions, and lexical variation
(e.g. ‘a very beautiful spot, a very beautiful part of this planet, of this world’) are
thus characteristic features of the interpreter’s rendering. However, the overall
impression for a reader (or analyst) of the English version is that both politicians
used a more informal style at the press conference. Moreover, the usually friendly
and informal way Bush employs when selecting journalists to ask their questions is
turned into a neutral question and answer format on the German website, as can
be seen in another extract from the 2006 press conference in Stralsund:
(14) President Bush: [. . .] President Assad needs to show some leadership

towards peace. [. . .]
Steve.
Q Thank you, sir. Just to follow up –
President Bush: Follow up on?
Q On both of these. Does it concern you [. . .] And on Iran, [. . .]
President Bush: I thought you were going to ask me about the pig.
Q I’m curious about that, too. (Laughter)
President Bush: The pig? I’ll tell you tomorrow after I eat it. The Iranian
issue is [. . .]
P Bush: [. . .] was den Frieden betrifft.

Frage: Um bei diesem Thema nachzuhaken: Machen Sie sich Sorgen
darüber, [. . .]
P Bush: Ich dachte, es würde um das Schwein gehen. [. . .] Die Iran-Frage
[. . .]
(Literally:
P Bush: [. . .] as concerns peace.
Question: To follow up on this topic. Does it concern you. . . [. . .]
P Bush: I thought you were asking about the pig, [. . .] The Iran issue [. . .])
We see that the specificity of the discursive practice of press conferences which
can be experienced by a reader of the White House website has become invisible
in the German version as a result of translation processes.
8.4 Conclusion
As the discussion of examples should have illustrated, translation cannot be limited

to a purely linguistic phenomenon. Translation competence involves more than
competence in two languages, it involves as well cultural, textual, domain-specific
competence (cf. Neubert, 2000). Translators operate within specific socio-cultural
settings; in short: translation is a socio-political practice. It is not the (linguistic
structure of the) source text which determines the linguistic structure of the target
text, but translators need to (be competent to) decide on the required structure
of the target text in each case in view of the purpose the target text is to fulfil for
its addressees in the target culture setting. This involves critical reflection about
socio-cultural, institutional, ideological aims, conditions and constraints. However,

before a translation gets published, other people have usually been involved as
well, e.g. revisors or editors. In some cases, texts are subject to censorship, with the
result that translations are brought in line with dominant ideologies at a particular
time in a particular culture (if translations are not banned from publication at all).
The development of the discipline of Translation Studies reflects this widening
of the perspective. Modern Translation Studies is not concerned with examining
whether a translation has been ‘faithful’ to its source text. Instead, the focus is on
social, cultural and communicative practices, on the cultural and ideological sig-
nificance of translating and of translations, on the external politics of translation,
on the relationship between translation behaviour and socio-cultural factors. In
other words, there is a general recognition of the complexity of the phenomenon
of translation, an increased concentration on social causation and human agency,
and a focus on effects rather than on text-internal linguistic features. It is also
widely accepted nowadays that Translation Studies is not a sub-discipline of applied
linguistics (or of comparative literature, cf. Bassnett and Lefevere, 1990: 12) but
indeed an independent discipline in its own right. However, since insights and
methods from various other disciplines are of relevance for studying all aspects of
translation as product and process, Translation Studies is often characterized as an
interdiscipline (cf. Snell-Hornby et al., 1992). In other words, with translation itself
being a crossroads of processes, products, functions, and agents, its description
and explanation calls for a comprehensive interdisciplinary approach.
One of the disciplines which contributes concepts and methods of analysis is
Applied Linguistics. Similarities and differences revealed as a result of a compara-
tive analysis of source texts and their target texts can be described with reference
to, among others, transitivity (e.g. shifts between passive and active voice), tenor
(e.g. shifts in the degree of formality), genre conventions (e.g. selection of saluta-
tion formula in letters). For explaining such translation strategies, additional con-
cepts from Critical Discourse Analysis (CDA) can be used as well. CDA analyses
discourse as a form of social practice, identifying and explaining the relationship
between a particular discursive event and the situation(s), institution(s) and social
structure(s) which frame it (cf. Fairclough and Wodak, 1997: 258). As I have argued
elsewhere (Schäffner, 2004), Translation Studies and Critical Discourse Analysis
share an interest in analysing (products of) human communicative activity in
socio-cultural settings. In this respect, processes of recontextualization and their
effects on linguistic-textual features can be mentioned. Any translation involves
recontextualization across linguistic, cultural and often ideological boundaries,
and as such involves transformations, such as deletion, rearrangement, substitu-
tion, addition. These transformations are translation strategies, and as is the case
in recontextualizations in monolingual settings, they are dependent on the goals,
values and interests of the context into which a specific discursive practice is being
recontextualized.
There is one final point which needs to be added: Politicians and journalists
engage with translations, they react to translated speeches, and/or quote extracts
in new texts. However, for some source texts there is not only one target text, but
several target texts. For example, the 2005 speech by the Iranian president Mahmoud
Ahmadinejad was translated into English by various institutions based in various
countries. The quote about Israel to be ‘wiped off the map’ which Tony Blair took
up in his critical reaction, mentioned at the beginning of this chapter, comes
from the translation done by the BBC Monitoring Service, for which the UK
government is among its main clients. In other translations, this particular phrase
had been rendered as ‘purged from the center of the Islamic world’ (http://memri.
org/bin/articles.cgi?Page=archives&Area=sd&ID=SP101305) or as ‘wipe this
scourge of shame from the Islamic world’ (http://www.iranfocus.com/modules/
news/article.php?storyid=4164). What is interesting to note is that one version was
taken up more frequently in subsequent reports than others. This illustrates the
power of some institutions and/or attitudes of politicians to texts as produced by
specific translation agencies – which makes any debate about accuracy of meaning
transfer irrelevant, or at least of secondary importance.
This example also points to limitations of Applied Linguistics for analysing
translations. A linguistic analysis and also a comparative analysis of different trans-
lations of the same source text can identify differences in the surface structures of
texts. A Translation Studies perspective can in addition explore the socio-political
and ideological conditions in which translations are produced, received and used
in subsequent contexts. Translation Studies can highlight sociocultural, political and
institutional practices, norms and constraints in the translation of political dis-
course and can thus yield insights into the intricacies of cross-cultural political
communication and interaction.
References
Baker, M. (2006), Translation and Conflict: A Narrative Account. New York and London:
Routledge.
Bassnett, S. (2004), ‘Trusting reporters: What exactly did Saddam say?’. The Linguist,
43, (6), 176–178.
Bassnett, S. and Lefevere, A. (eds) (1990), Translation, History and Culture. London:
Pinter.
Bell, A. (1984),‘Language style as audience design’, Language in Society, 13, 145–204.
Bielsa, E. (2007), ‘Translation in global news agencies’, Target, 19, (1), 135–155.
Blackledge, A. (2005), Discourse and Power in a Multilingual World. (Discourse Approaches
to Politics, Society and Culture 15). Amsterdam/Philadelphia: John Benjamins.
Burkhardt, A. (1996), ‘Politolinguistik. Versuch einer Ortsbestimmung’, in J. Klein
and H. Diekmannshenke (eds), Sprachstrategien und Dialogblockaden. Linguistische
und politikwissenschaftliche Studien zur politischen Kommunikation. (Sprache, Politik,
Öffentlichkeit, vol. 7). Berlin and New York: de Gruyter, pp. 75–100.
Calzada Pérez, Ma. (2001), ‘A three-level methodology for descriptive-explanatory
Translation Studies’, Target, 13, (2), 203–239.
Chesterman, A. (1997), Memes of Translation. Amsterdam and Philadelphia: John
Benjamins.
Chesterman, A. (1998), ‘Causes, translations, effects’, Target, 10, (2), 201–230.

Chilton, P. and Schäffner, C. 1(997), ‘Discourse and politics’, in T. van Dijk (ed.),
Discourse Studies: A Multidisciplinary Introduction, vol. 2: Discourse as Social Interaction.
London: Sage, pp. 206–230.
Fairclough, N. and Wodak, R. (1997), ‘Critical discourse analysis’, in T. van Dijk (ed.),
Discourse Studies: A Multidisciplinary Introduction, vol. 2: Discourse as Social Interaction.
London: Sage, pp. 258–284.
Holland, R. (2006), ‘Language(s) in the global news: Translation, audience design and
discourse (mis)representation’, Target, 18, (2), 229–259.
Kang, J.-H. (2007), ‘Recontextualization of news discourse: A case study of translation
of news discourse on North Korea’, The Translator, 13, (2), 219–242.
Koskinen, K. (2000), ‘Institutional illusions: Translating in the EU Commission’, The
Translator, 6, (1), 49–65.
Koskinen, K. (2001), ‘How to research EU translation?’, Perspectives, 9, (4), 293–300.
Lilleker, D. (2006), Key Concepts in Political Communication. London: Sage.
Mason, I. (2000), ‘Audience design in translating’, The Translator, 6, (1), 1–22.
Neubert, A. (2000), ‘Competence in language, in languages, and in translation’, in
C. Schäffner and B. Adab (eds), Developing Translation Competence. Amsterdam and
Philadelphia: John Benjamins, pp. 3–18.
Schäffner, C. (2004), ‘Political discourse analysis from the point of view of translation
studies’, Journal of Language and Politics, 3, (1), 117–150.
Schäffner, C. (2005), ‘Bringing a German voice to English-speaking readers: Spiegel
International’, Language and Intercultural Communication, 5, (2), 154–167.
Schäffner, C. (2007), ‘Politics and translation’, in P. Kuhiwczak and K. Littau (eds),
A Companion to Translation Studies. Clevedon: Multilingual Matters, pp. 134–147.
Seymour, E. (2002), ‘A common EU legal language?’, Perspectives 10, (1), 7–14.
Simon, S. (1996), Gender in Translation: Cultural Identity and the Politics of Transmission.
London: Routledge.
Snell-Hornby, M., Pöchhacker, F. and Kaindl, K. (eds), (1992), Translation Studies.
An Interdiscipline. Amsterdam and Philadelphia: John Benjamins.
Tosi, A. (ed.) (2002), Crossing Barriers and Bridging Cultures. The Challenges of Multilin-
gual Translation for the European Union. Clevedon: Multilingual Matters.
Toury, G. (1995), Descriptive Translation Studies and Beyond. Amsterdam: John Benjamins.
Tymoczko, M. and Gentzler, E. (eds) (2002), Translation and Power. Amherst: University
of Massachusetts Press.
Wagner, E., Bech, S. and Martínez, J. M. (2002), Translating for the European Union Insti-
tutions. Manchester: St Jerome.
CHAPTER
9
Language and the Law
John Gibbons
9.1 Introduction
In Gibbons (2001) I suggested that the applied linguistic enterprise has three main
stages – a revealing and analysis of a language problem or issue (reflection), the
development of some form of treatment (action) and evaluation of the success of
the treatment. This approach frames the following discussion of Language and
the Law.
Since this chapter is in English, I shall talk mostly of the Common Law system.
It is important to note however that many more people are subject to versions
of the Roman Law system, in most of Asia (including China, South-East Asia and
Japan), Latin America and continental Europe, and that Shariah law is also
widespread.
9.2 Legal Language
The language of the law is a significant concern, because the law is such an impor-
tant and influential institution, and because it is packed with language problems.
Most of our common everyday activities are carried out within a legal frame. A bus
ticket is a legal contract, and virtually any form of transport, particularly driving a
car, is similarly hedged about with legal issues. Employment too is a legislative
domain. Our family relations are subject to family law, and the media we use are
similarly controlled. Law intrudes into almost every aspect of modern life. If we
examine the law however, it is the most linguistic of institutions. Legislation is a
linguistic entity, with no existence outside of language. Equally police investigation
and court proceedings are overwhelmingly linguistic processes, mainly spoken
rather than written in Common Law systems. What makes these of significance for
applied linguists is that the linguistic aspects of the law raise many issues and diffi-
culties. The written language of legislation and regulation is difficult to under-
stand for lay people – there is a profound communication problem – yet lack of
understanding of the law is not a defence in court.
To address these real-world issues we need to understand the nature of legal
language, and possible sources of communication difficulty (the reflection stage),
and to work out ways of overcoming the problem in so far as this is possible (action).
Language and the Law 165
The analysis here examines four major sources of possible problems. The first is
the specialized text structures and procedures used in the law – the genre issue in
short. The second is the extreme writteness of many legal documents: some are
virtually impossible to read aloud in a meaningful way. The third is the technicality
of much legal discourse: the law and its practitioners have developed a range of
unique legal concepts, and these can be expressed efficiently only by using legal
jargon. The fourth is the interpersonal arena, where power disparities and hyper-
formality are produced by the essentially controlling nature of the legal system.
Looking first at genres, the highly institutionalized, and sometimes ritualized
discourse of the law often follows regular patterns: organized sequences of ele-
ments which each play a role in achieving the purpose of the discourse. Following
V. K. Bhatia (1993) Halliday and Hasan (1985) and Martin (1992) among others,
these are termed genres. It is well established in reading theory that a knowledge
of the genre that one is reading is important, and sometimes essential for under-
standing (Wallace, 1990; Weaver, 1988). This is in part why legal documents can
be difficult for lay readers to understand, while lawyers have less difficulty. This is
well illustrated in the discussion between the eminent linguist Charles Fillmore,
and some legal authorities reported in Washington University Law Quarterly,
Volume 73 (1995: 922–931), particularly the discussion of the following sentence
from a contract.
After this marriage in the absence of any agreement to the contrary the legal
relations and powers as regards to property might, by reason of some change in
our domicile or otherwise, be other than those of our present domiciles or other
than those which we desire to apply to our relationship powers and capacities.
Charles Fillmore finds this ‘incompetent’ and unintelligible. The lawyers, how-
ever, were able to draw on their knowledge of the genre of contracts of this type,
to say that such a clause is inserted at the beginning of many such contracts to
cover the contingency of the parties moving to another state where the law is
different. Despite the chronically poor drafting of this language, their knowledge
of the genre enabled the lawyers to understand it and to be in agreement concern-
ing its meaning.
One of the most fundamental genres is that of narrative. An important issue
arising from genres is the notion developed in some depth by Bennett and Feldman
(1981), Jackson (1991) and Stygall (1994) that the competing (prosecution or
plaintiff vs. defence) versions of events in a trial are in fact competing narratives.
Bennett and Feldman (1981) describe these as competing ‘stories’. Courtroom
narratives are not limited to the particular events under litigation, they may be
stories of the witness’s life, loves and previous contacts with the law. This narrative
interpretation both facilitates analysis, and problematizes the process. For exam-
ple, narratives mostly follow a simple linear time sequence, yet life is rarely simple
or linear – events happen at the same time and relationships between them may be
subtle and complex. However, some of the authors in Papke (1991) provide evi-
dence from trials that a simple narrative structure easily intelligible to jurors may
be preferred to a more complex account that is closer to the facts. Again the risks
of injustice inherent in such language behaviour are troubling.
166 John Gibbons
Courtroom proceedings, and police procedures can also be seen to follow genre
structures. Indeed in the case of trials, the sequence of stages through which they
pass is regulated. Maley (1994: 16) provides a helpful chart listing the main genres
used in the legal process. Once more a knowledge of these genres is helpful in
understanding and participating in what is happening. Hall (2004) makes a con-
vincing case that the possible different purposes of police interviews, seeking the
truth of events or attempting merely to get a result by means of a confession,
demand different genre structures, and the current prescribed genre structure of
police interviews in New South Wales inadvertently favours the second. Rock (2007:
chapter 5) working in the UK context, shows how the genre structure of the Anglo-
Welsh police caution can be a barrier to communication. She also shows how
important it is to see oral cautioning as a social process. In this case the form of the
Applied Linguistic treatment is obvious: change the genre and train police officers
in its use.
Turning now to the written-spoken dimension, it is worth remembering that the
all legal systems have oral origins – the Roman legal system, the source of most
continental and Asian legal systems, was an oral system for most of the existence of
the Roman empire; the Common Law system used in the English speaking world
has its origins in Germanic tribal law; Shariah, the Islamic legal system, developed
in part from the orate systems of desert Arabs – indeed the Prophet Mohammed
(PBUH) was probably illiterate. Danet and Bogoch (1994) provide a convincing
account of the movement from an oral to a written mode. The linguistic conse-
quences are far-reaching. They include the possibility of extremely long and
complex sentence structures (often between 70 and 100 words), which are virtu-
ally impossible to read aloud meaningfully or to understand when heard. With
written language however we have the luxury of multiple recasts to construct them,
and multiple readings to decode them, so that they become possible, if perhaps
undesirable. These very long sentences are often used to package together a
number of core concepts or prescriptions, along with all the conditions in which
they apply (V. Bhatia, 1994). Legal language also tends to use long and complex
noun phrases; Crystal and Davy (1969: 205) give the following example: ‘The
payment to the owner of the total amount of any instalment then remaining
unpaid of the rent herinbefore reserved and agreed to be paid during the term . . .’.
Halliday (1989) and Halliday and Hasan (1985) show that this process, particularly
the creation of abstract nouns from other parts of speech (‘grammatical meta-
phor’) is a consequence of literacy. An illustration of these phenomena can be
found in the short extract from a contract given previously. It is a sentence of
55 words, with a fairly complex structure including a number of prepositional
phrases and a long complex phrase following ‘be’. There are also numerous
abstract nouns, including ‘absence’, ‘agreement’ and ‘capacities’. Police investiga-
tion and court proceedings can also suffer from this problem, although to a lesser
degree, since oral interaction includes the possibility of the face-to-face negotia-
tion of meaning. The problem is that there is psycholinguistic evidence (Felker
et al., 1981) that complex sentence and phrasal structures, and the use of gram-
matical metaphor, make texts difficult to understand.
Another part of the difficulty of legal language is its technicality. Maley (1994)
among others has pointed out that the law consists to some degree of legal con-
cepts, and therefore words to express these concepts are essential. It is part of a
wider objective of legal language, that of being as precise and decontextualized as
possible. Some legal terms are used almost exclusively to refer to legal contexts, for
instance ‘estoppel’ and ‘magistrate’. Others are words with non-legal meanings
that are used with a particular meaning in legal contexts, such as ‘party’ (one side
in a court case), ‘damages’ and ‘restraint’ (as in ‘restraint of trade’). Legal English
has borrowed a range of terms from Norman French and Latin, and many of these
terms are still in wide use: for example habeas corpus and voir dire. Jargon is also
characteristic of the legal language of other European languages, and to a lesser
extent of legal Japanese. Legal Chinese on the other hand uses mostly everyday
language with specialist meaning. In general, laws and contracts are intended to
apply to specific behaviours and entities/people in specific circumstances, and
legal language attempts to spell out precisely what these are, in order to avoid hos-
tile interpretations. Legal drafters often combine all these elements in a single
sentence, which explains the extreme length mentioned earlier.
The language problem that arises from technicality is that legal terms limited to
the legal domain may not to be know to lay people, and legal terms with non-legal
meanings may be understood in their everyday sense. For example Diamond and
Levi (1996: 232) mention jurors misunderstanding the legal term ‘aggravating’ to
mean ‘irritating’, and thereby being in danger of incorrectly imposing the death
penalty. If legal jargon is in some cases unknown to or poorly understood by non-
lawyers, it clearly has the potential to impair their understanding of and their par-
ticipation in the legal process. This accentuates the problems of complexity
discussed above.
One source of this difficulty is that legal discourse may be addressing two audi-
ences, both a lay audience and a legal audience. For instance police cautions must
not only communicate to the person being cautioned, they must also be admissible
in court as having fully performed the function of cautioning. This explains in part
inertia and even resistance when it comes using plain language for legal purposes.
Another source of resistance among police and lawyers is their understanding of
the types of social message conveyed. Most work in this area has assumed that only
propositional information is communicated. It is clear however that complex and
technical language also carries a social message concerning the power and author-
ity of the person using it. Resistance from lawyers and police officers to a lessening
of this power and authority is not surprising.
One ‘action’ taken to resolve this problem is the adoption of plain language
principles, which attempt to make the language of the law as simple and compre-
hensible as possible, while ensuring that the legal language continues to perform
its task of being as explicit and watertight as possible (see for example Law Reform
Commission of Victoria, 1987; Steinberg, 1991). There are active plain legal lan-
guage movements in the United States, Britain and Australia, and in a range of
other countries. In Britain they have recently achieved a significant success in per-
suading the government and mainstream English law authorities to implement
168 John Gibbons
change towards plain language. They have assisted the Master of the Rolls Lord
Woolf in producing the 1998 Civil Procedure Rules (SI 1998 3132) which substi-
tute many forms, documents and procedural wordings with clearer equivalents.
Many of the idiosyncrasies of legal language here have been addressed. A number
of arcane legal terms have been replaced: for instance a plaintiff is now a claimant,
a pleading is now a statement of case. Law Latin has been replaced with English –
ex parte, inter partes, in camera and subpoena have become with notice, without
notice, in private and a summons. Proper names such as an Anton Piller order
have been replaced with more transparent titles such as a search order. These
changes have yet to be adopted elsewhere, and critical evaluations of them are
emerging.
However, the editing involved in the process can carry risks, as Davies (2004)
shows: in the removalist’s contract that she discusses some of the legal content is
lost in a plain language version. Gibbons (2001) demonstrates some of the prob-
lems caused by the linguistic complexity of police cautions, discusses the some-
times tortuous process of simplifying them and provides evaluations of the revisions.
Rock (2007: chapter 13) suggests more wide-reaching action, by including within
the legal process specific checking interactions, and individually crafted responses
on the basis of these checks which target individual comprehension and participa-
tion issues.
When we examine interaction in legal contexts, another issue arises – extreme
power asymmetry. The legal system is by its very nature an instrument of control
and power, and in a democratic society this power is ceded to the legal system to
maintain order and some degree of fairness within society. However, the power
ceded to police and lawyers runs a constant risk that people will be coerced into
saying things they do not mean or know to be untrue or incomplete. This interper-
sonal power is manifested and exerted to a significant degree through language.
It can be seen in the forms of address used: Your Honour, Your Worship, Your
Lordship, My Learned Friend, etc. Among police officers it may lead to the use of
unnecessarily elaborate ‘copspeak’ to maintain status: Maley and Fahey (1991: 8)
give the following example from a police sergeant’s courtroom testimony.
Police Officer: I was unable to maintain the light being illuminated.

Counsel: To keep the torch on?
Police Officer: To keep the torch on.
Power relations in both courtrooms and police stations affect turn taking. In a
court it is illegal for people to speak without being allotted a turn by the judge, and
illegal for them not to speak when questioned, unless they have a specific ‘right to
silence’. While judges have the right to speak whenever they wish, lawyers in gen-
eral must take turns (Atkinson and Drew, 1979). Turn taking in courtrooms has
therefore become regulated and institutionalized along power hierarchy lines.
Equally, police officers when interviewing will often refuse to answer questions
and will expect answers. The coercive nature of courtroom questioning has also
received considerable attention from linguists – see for example Danet et al. (1980),
Harris (1984), Phillips (1987). Cotterill (2003) shows how power was manipulated
strategically in the O. J. Simpson trial. Eades (1994) gives the following example
of highly coercive questioning when a young Aboriginal witness remained silent
during cross-examination
Counsel: . . . I’d suggest the reason to you, because you don’t want everyone to
know the little criminal that you are, do you? That’s the reason, isn’t
it? Isn’t it? Your silence probably answers it, but I’ll have an answer
from you. That’s the reason, isn’t it?
The core information ‘you are a little criminal’ is deeply embedded in the gram-
matical structure, so that it is very difficult to deny directly – a negative response
would be a denial that this is the reason for silence, not a denial that he is a
criminal. Furthermore there are multiple question tags such as ‘isn’t it’ used as
coercive devices. Eades argues that the content of any answer to such a question
would be suspect. Lawyers use many such linguistic strategies to control the
responses of witnesses.
Heffer (2006) also shows other characteristics of courtroom discourse, involving
legal professionals and lay people. For example he shows how a judge’s summing
up can influence jurors.
Critical discourse analysis is an emerging focus. For instance Vasilachis de
Gialdino (1997) examined an Argentinian labour reform bill rooted in neo-
liberalism, describing the language used within labour courts in Argentina, discus-
sion of the reform in the parliament and the executive, and the treatment of these
in the local press, showing in the latter (Vasilachis de Gialdino, 1997: 270–271)
that workers were not discussed, unionists were portrayed as violent and irrational,
and reduced protection for workers was portrayed as a positive move towards
globalization, modernization and flexibility. There is also a growing debate con-
cerning gender and language in the law, often showing an interaction between
legal power and male-female power relations (see particularly Matoesian, 1997).
This is related to language and disadvantage before the law (see below).
The problem is clear, in that truth may be the casualty when questioning
takes place in situations of high power asymmetry in which the witness is open to
manipulation. It is difficult to find thoroughgoing solutions. Some actions taken
so far are to alert and educate lawyers and judges to the risks involved in such ques-
tioning, to change the rules of courtroom procedure to reduce the use of coercive
questioning, and where particularly vulnerable witnesses are involved (such as
children, the intellectually handicapped, and the deaf) to allow the presence of
a ‘friend’ to support and help them. However, the problem is deeply rooted in the
adversarial nature of Common Law legal systems, and the notion that evidence
must be ‘tested’.
The language of witnesses may also manifest power or its absence. O’Barr (1982)
and O’Barr and Conley (1990) did an important series of studies which revealed a
set of linguistic markers of power (such as hesitation, low coherence and use of
emphatics and mitigators), and demonstrated that witnesses and defendants whose
170 John Gibbons
language is less powerful were less likely to be believed – a worrying indicator of

the linguistic means by which social injustice may be reproduced. There were even
indications that people who use less powerful language might receive less financial
compensation from an offender.
The nature of the language of the law poses other applied linguistics challenges.
How can it be taught? How can it be translated or interpreted into other
language?
9.3 Teaching the Language of the Law
We have seen in previous sections the extreme complexity and unusual nature of
legal language. This poses a substantial problem, particularly for the many coun-
tries where the language of the law is not the mother tongue of those involved in
the legal system. In India and much of anglophone Africa for example, lawyers in
training need help to master not only technicalities and the legal concepts that
they represent, but also the convoluted grammatical structures in which much leg-
islation is framed. This places considerable demands upon the teachers and cur-
riculum designers responsible for teaching English to these law students. They
themselves may have trouble in understanding the cognitive complexity of legal
documents and the linguistic realization of that complexity. Once understood,
training students to master it is a pedagogical challenge. Teachers may also need
to train law students in oral interactive techniques to master the power laden lan-
guage of the court. There are also ethical issues involved in both the promulgation
of this register that excludes so many ordinary people, particularly second language
speakers, and in training people to use language to manipulate and distort the
testimony of others. For a more extensive survey, see Northcott (2008).
9.4 Legal Interpreting and Translation
Turning first to legal interpreting and translation, we have already touched on the
possible disadvantage suffered by minorities who cannot cope easily with the com-
plexity and power of the language of the law. People who have only a limited com-
mand of the language used for legal proceedings are also likely to suffer severe
disadvantage before the law if (a) an interpreter/translator is not provided and
(b) if the interpreting/translation does not accurately convey what is said/written.
Although most legislations will provide such services in some circumstances, the
basic Common Law situation is that it is at the discretion of the judge. Since judges
are rarely qualified language testers, there is chronic under provision of interpret-
ers in some jurisdictions. For instance Carroll (1995) describes some courts
where interpreters are provided in less than 10 per cent of the cases where they
are needed. Gibbons (2001) describes a similar situation for police under-use of
interpreters. Some jurisdictions (such as the state of New Jersey in the United
States) have adopted laws which address this issue, by making interpreters availa-
ble for all second language speakers unless there is evidence from a qualified
language tester that the person has sufficient command of the courtroom lan-
guage to fully participate in proceedings.
There may also be a problem with interpreter supply, particularly for languages
of low demand, or where legal interpreter training is unavailable. The paradigm
case is that of tribal minorities whose language has a small number of speakers,
where there may be no highly proficient bilinguals, or no appropriate interpreter
training available in the community. Medium-term solutions include the training
of para-professional interpreters to provide at least some service to the community,
and long-term solutions will involve investment in education, and the develop-
ment of alternatives such as minority language courtrooms.
Berk-Seligson (1990), Hale and Gibbons (1999) and many others have docu-
mented the extreme difficulty of providing accurate interpreting in courtroom
contexts, where even minor inaccuracies may lower the standards of justice. The
conditions that make this process particularly difficult are the pressure to use as
little time as possible (by its very nature interpreted testimony takes twice as long),
and the lack of understanding of interpreting among some lawyers, who may
for example interrupt during interpreting, or demand a literal word-for-word
translation. In order to avoid such problems interpreters are often reluctant to use
dictionaries, or to ask what is meant when there are two candidate translations.
All these factors can reduce the accuracy of interpreting. Particular linguistic
problems include the interpreting of address forms (e.g. señor in Spanish), passivi-
zation, discourse markers such as ‘umm’, ‘well’, ‘you know’, and tag questions
(there are no exact equivalents of English tags in other languages, but as we saw
in the example from Eades given above, they are an important feature of cross-
examination). In many jurisdictions these problems can be exacerbated by inade-
quate training and sub-professional rates of pay. Clearly adequate resourcing is
a basic first step in resolving this issue, with more training for both lawyers and
interpreters, but the very nature of interpreting and translation is that it is not an
exact process – a consequence of the differences between human languages and
cultures.
Turning now to legal translation, the problems do not lie with the interactive
phenomena discussed in relation to interpreting, but rather in the extreme com-
plexity and technicality discussed earlier. For example Vlachopoulos (2004) dis-
cusses the translation of an English language legal document into Greek, and
documents the challenge posed by Common Law concepts (and the terms used to
refer to them), which in a number of cases do not exist in Greece’s continental
legal system, and therefore lack a corresponding Greek term. When we add in the
conceptual complexity and delicacy of many legal texts, the task becomes even
more daunting. Vlachopoulos proposes a range of solutions including the use of
terms which are close in conceptual content from non-legal registers including
everyday language. Otherwise one is obliged to use extensive footnoting and dis-
cussion of the translation itself.
Apart from the language of the law, there are a number of other areas where
language issues emerge in the legal arena. Important among these are language
crimes, language legislation and linguistic evidence.
172 John Gibbons
9.5 Language Legislation
There is legislation on many language issues. One area is that of language rights.
A language issue that underlies many armed conflicts around the world (for
instance Macedonian in Kosovo, or Kurdish) is the right to use a language for pub-
lic purposes such as education, law and with government agencies, and even to
speak it in private. There has been a movement in the European Union towards
the acceptance of many more languages as public languages – for instance Catalan
and Basque in Spain. An indication of how far this has gone is that judges in the
Basque country must learn Euskara, or else employ an Euskara interpreter at their
own expense. In the United States on the other hand most states now have legisla-
tion to prevent the use of minority languages for public purposes (González and
Melis, 2001). The basic argument seems to be between the role of the dominant
language in sustaining national unity and including all members of society in its
processes, and the rights of minorities to access public institutions in a language
they fully understand, and to maintain their language and culture. Proponents of
US English do not seem to take sufficiently into account the evidence that for chil-
dren a high level of bilingualism is viable and achievable, while for adult migrant
learners of the second language high levels of proficiency are rarely attained, and
therefore services in their mother tongue are needed to avoid social disadvantage.
The actions taken to support or suppress the use of particular languages in national
life is the passing and enforcing of legislation or regulations. The evaluation of
their success can be seen in long-term language shift, maintenance and loss.
There is also a type of language legislation by means of which certain language
behaviour is criminalized to become ‘language crimes’. Examples are bribery,
threats and perjury. Shuy (1993) provides a thorough analysis of the linguistic
nature of such crimes and also reveals the difficulty and delicacy of demonstrating
in court that such crimes have or have not been committed. For example, he shows
that for the successful achievement of bribery there is a genre consisting of an
opening; a discussion of the briber’s ‘problem’ and checking that the bribee has
the capacity to intervene in the problem; a proposal, which may involve some
negotiation of both action and reward; an acceptance or rejection of the bribe; if
the bribe is accepted the possible discussion of future ‘business’; and a closure. For
bribery to take place both the proposal and the acceptance are essential stages.
Shuy shows that it is not uncommon for cases to come to court in which it is clear
that the bribee did not accept the bribe but is being prosecuted for being part
of a bribery event. His painstaking analysis is a prerequisite for action, in this
case appearing in court and showing as appropriate that the language crime of
bribery did or did not take place. Green (1990) documents a case where a young
man was accused of conspiracy to distribute cocaine. The question in this case
was whether the young man participated in a drug deal. Green shows through a
careful and detailed discourse analysis of pronoun use, and the man’s contribu-
tions to the discussion, including markers of cooperation (such as ‘yeah’), answers
to questions, topic management, clarification requests, interruptions, turn taking,
as well as incomprehension markers, that he never actively participated in the deal
itself. The evaluation of the success of such intervention by forensic linguists is usu-
ally whether their evidence is accepted by the court, and is affirmed if they are also
on the winning side.
Another type of language crime is that of using offensive language, mostly
swearwords. For example, the NSW 1988 Summary Offences Act states:
(1) A person shall not -

...
(b) use offensive language in or near, or within hearing from, a public place or
a school.
(2) It is sufficient defence to a prosecution for an offence under this section if
the defendant satisfies the court that the defendant had a reasonable excuse
for conducting himself or herself in the manner alleged in the information
for the offence.
Previously the maximum punishment of this offence was up to three months in

prison, changed to a fine in 1993. The question asked by B. Walsh (1995) is ‘What
is offensive?’ The test proposed by the courts is ‘whether reasonable persons in the
relevant place and at the relevant time, and in the circumstances there and then
prevailing, would be likely to be seriously alarmed or seriously affronted’. It is
noticeable that this test depends on the immediate context, including the partici-
pants and their schemas. It provides little real information, leaving it open to the
magistrate to determine who is a reasonable person, and what is likely to alarm or
affront them. There is also a defence of a ‘reasonable excuse’, for instance if some-
one drops a hammer on their foot. On reflection, the extreme discretion within
this legislation is dangerous, since in 1993 Amnesty International reported that it
was used overwhelmingly to imprison or fine Aborigines, and to a lesser extent
younger and working class people. Around 5,000 people a year were found guilty
of this offence in the mid-1990s. The applied linguistic action adopted by Taylor
(1995) reveals the hypocrisy of law makers and law enforcers, by gathering well-
documented instances of police officers and politicians using swearwords them-
selves. Indeed one aboriginal man was prosecuted for saying to police ‘Don’t tell
me to get fucked’. Even if one deplores the use of offensive language, in these cir-
cumstances to legislate against it is unfair and unrealistic and questionable given
the history of such legislation in oppressing minorities.
Another type of language crime is vilification. This can take the form of libel or
slander of individuals, or of group vilification. To be prosecuted, libel and slander
of individuals need evidence of untruth and of harm to the recipient, but in law
there may be no need to prove intent, i.e. that the harm was deliberate. So the defi-
nitions of slander in normal dictionaries normally include intent, for instance the
Oxford Dictionary definition has ‘maliciously’, while law dictionaries exclude this
element: another case where legal constructions differ from those of everyday
language and culture. Interestingly, in the United States freedom of speech con-
siderations have taken precedence, and there is little litigation concerning libel
and slander, at least in cases involving the media. However, in other Common Law
174 John Gibbons
jurisdictions where freedom of speech is not constitutionally guaranteed there are

many more court cases.
Group vilification is mostly legislated against in terms of ethnicity, but there is
also legislation against vilification on the grounds of religion, disability or sexual
orientation. Group vilification usually takes the path of constructing an ‘us’ and a
‘them’ (often on little real basis), and then negatively portraying ‘them’ (see for
example van Dijk, 1987, 1993). The problem is that vilification, as well as causing
distress to its recipients, can lead to discrimination and even violent action. The
Nazis for instance consistently portrayed the Jews in words and image as vermin,
and this served as a rationale for extermination. However, the legislative action to
be taken is hotly debated, with the United States largely refusing to inhibit freedom
of speech, while some other legislations do so (Freedman and Freedman, 1995).
Evaluation of the success of such legislation has shown that it tends to lead to
coded expression of vilification in place of overt expression.
9.6 Linguistic Evidence
Finally, another major area where language intersects with the law is that of linguis-
tic evidence. The paradigm case is where a linguist or applied linguist gives evidence
on a language issue in court. However, such evidence may also be provided to police,
lawyers and intelligence agencies, and in a range of other contexts. Various types of
expert and expertise may also be involved, including anthropologists and sound
technicians.
One issue that must be addressed is the admissibility of linguistic evidence. In
Roman Law based legal systems there is usually a system for accrediting experts.
After rigorous examination, an expert is accepted or rejected as competent to give
evidence in a certain field, and if accepted, his/her evidence will be taken without
further demur in subsequent court cases. In Common Law systems the compe-
tence of expert witnesses is challenged each time they appear in a case, and their
evidence is accepted or rejected on the following grounds: expertise – whether
their knowledge is specialized and beyond ‘common sense’ knowledge; validity –
whether their expertise and evidence is fully relevant to the issue on which they are
testifying; and reliability – whether it is scientifically derived. Bowe and Storey
(1995: 188–189) point out concerning the expertise of forensic linguists:
While many people are quite capable of identifying or eliminating unknown

speakers in a[n earwitness] line-up, they are generally unable to say why . . .
Linguistically trained analysts on the other hand are in a position to give a detailed
description of differences or similarities noted in two voice samples, together
with an explanation of how and why these differences or similarities occur.
Evidence may range across many linguistic levels, including phonology, grammar,
discourse and conversational phenomena, and sociolinguistic variation. Linguistic
evidence falls into two main areas – communication and identification. We will
examine these in turn, beginning with communication, and moving through the
linguistic levels.
9.6.1 Communication
Looking first at the role of pronunciation in communication, a linguist may be

called upon to uncover what someone said. For example I have been involved in
two cases where some form of secret language or ‘pig Latin’ was used, where my
role was to decode it. This may not always be as simple as one might imagine, for
example in one case, during early hearings of a tape recording it was difficult to
crack the oral code used in expressions such as [bəpəkəpos əpin əpu kəpapəpu:
əpov məpunθs]. It emerged that every vowel has an [əp] inserted before it (it
reads ‘because in a couple of months’). Linguists may also be called upon to say
whether an accent or a poor quality recording causes intelligibility problems.
At the level of vocabulary and grammar, linguists may be able to say both what is
meant by a particular wording, or whether particular complex lexical and/or
grammatical forms make a text difficult to understand. Levi (1994: 16–17) dis-
cusses her evidence in a case where the information given to recipients of ‘public
aid’ was done in language that was virtually unintelligible to them. She writes:
my analysis included commentary on such problems as use of bureaucratic jar-

gon, crucial terms left vague or undefined, needlessly complex syntax, anaphora
(eg. demonstrative pronouns like this) with obscure antecedents, related infor-
mation scattered throughout nonadjacent sections, incoherent sequencing of
topics, blatant omissions of critical (and legally-mandated) information, and an
intimidating and obfuscating graphic presentation.
From this it can be seen that Levi also examined discourse phenomena. The
success of Levi’s intervention can be seen in the fact that the agency involved was
ordered to pay US$20 million to the recipients, and to rewrite its documents in a
way intelligible to them.
McMenamin (1993) documents a case where the issue was the meaning of the
words ‘syndrome’, ‘accident’ and ‘disease’. McMenamin testified on behalf of parents
whose child died of Sudden Infant Death Syndrome (SIDS) at the age of 18 months.
The child’s life was protected by the father’s life and accident insurance, which
included the statement ‘The plan pays a benefit for losses resulting from any kind
of accident . . .’. The insurance company denied the claim initially, saying that
the policy did not cover deaths from illness or disease. McMeniman’s reading of
the medical literature and dictionary definitions revealed that ‘syndrome’ is distin-
guished from ‘disease’ in that a ‘syndrome’ groups together patterns of incidents,
but there is no explanation in terms of physical malfunctioning, particularly that
caused by bacteria, viruses etc., while ‘diseases’ exist at a specific time in a specific
person, between health and either restored health or death (‘syndromes’ do not
share this quality). As McMeniman says
‘disease’ is a temporally bound state between health and death. A diseased per-
son either gets well or becomes chronically diseased and dies. SIDS is something
a healthy infant either has or does not have. The result, even with a ‘near-miss’,
is health or death, nothing in between.
176 John Gibbons
Hence SIDS cannot be classified as a disease. McMeniman’s evidence was accepted.

Solan (1995) makes a case that linguists could be involved in decisions concerning
the meaning and application of legislation in particular cases (legal interpreta-
tion) on the basis of grammatical and lexical analysis, but this is challenged by
lawyers (see the debate in the special edition of the Washington Law Journal in
which Solan’s paper appears).
Over the last 20 years Diana Eades and Michael Walsh among others have
carefully documented communication problems between Australian Aborigines
and the law (see for instance Eades, 1994, 1995, 2000, 2008; M. Walsh, 1994).
They describe the problems that speakers of Aboriginal English have with the legal
process. An underlying issue is what is known as the ‘knowledge economy’ in
Aboriginal society. In traditional Aboriginal societies material goods were mostly
held in common, and status rather than deriving from wealth came from the
possession of secret knowledge (this situation is also found in other indigenous
communities). The result is that much knowledge is not to be shared freely. Some
of it is available only to those who have been ceremonially initiated into it. It may
be the property of only women or men (women’s/men’s ‘business’). The conse-
quence is that questioning in Aboriginal societies is generally done with great
caution, often indirectly by raising a topic, and leaving it open to the interlocutor
to contribute what knowledge she/he is willing to share. Direct questioning is
regarded as rude and intrusive. Answering is not obligatory, since a direct answer
may involve secret material or may grant the questioner unearned status. The
clash with police questioning and courtroom examination is evident. Police inves-
tigation and court trials are largely dependent on the right to question and the
obligation to answer. Eades has testified in court on a number of occasions con-
cerning the resulting lack of communication.
9.6.2 Identification
Identification may involve comparing two or more language samples, and saying
whether they were produced by the same person or not; alternatively it may involve
profiling the person who produced the language – there may be indicators of age,
class, occupation, gender, mother tongue and so on. Perhaps the best known area
of identification is that of speech sounds – there are many papers on this topic
in the journal Forensic Linguistics, and Hollien (1990, 2001) provides detailed and
convincing description and illustration of the issues involved. Perhaps the most
tendentious issue is whether machine analysis is superior to the expert ear. There
was a period when ‘voiceprints’ (more correctly spectograms) were widely used in
the United States, often by people poorly trained to produce and interpret them.
Not surprisingly much of this evidence was discredited subsequently, which led to
considerable suspicion of such methods in courtrooms. Hollien (1990) provides
spectograms of a particular utterance, where two different speakers had almost
identical spectographic profiles, and a single speaker produced markedly different
profiles. In recent years the techniques and technology have developed, and much
more caution is used in drawing conclusions. In particular certain vowel formants
can be a strong contributor to voice identification, provided that the recording on
which they are based is of adequate quality. Hollien notes, however, that there are
many points where recording quality can be lost. Identification becomes particu-
larly difficult if one of the samples for analysis is recorded in such a way that much
of the signal is affected, for example over the telephone, or on poor recording
equipment, or in a poor recording situation (for instance one involving back-
ground noise): since many police recordings of voices are covert, it is unlikely that
these will be of high quality. Sometimes therefore the human ear is better indica-
tor than a machine, particularly when, for example, one is attempting to distin-
guish between regional accents (in our current state of knowledge this task cannot
be performed by a machine). Often a combination of the two techniques is effec-
tive. Courts still tend to prefer machine based analysis, since it is more overtly
‘scientific’. The untrained ear is unreliable in voice identification, earwitness iden-
tification being even less reliable than eyewitness identification.
Where speech sound data may be reliable is in the negative: it is often possible
to say with certainty that two samples come from different speakers, even if it is not
always possible to say with certainty that two speech samples come from the same
speaker. Labov and Harris (1994) describe the Prinzivalli case, in which Labov says
that there was no doubt that Prinzivalli could not have made a bomb threat phone
call because the bomb threat voice had an unmistakable New England (Boston
area) accent, while Prinzivalli had an equally unmistakable New York accent. His
main problem was convincing the court of this (see Labov and Harris for a clear
exposition); in essence he had to train the court to hear the pronunciation differ-
ences. The evidence was accepted and Prinzivalli was acquitted.
Similar in nature are various identificatory elements of the written language.
Handwriting can be distinctive, and peculiarities of spelling and punctuation can
be strong identifiers. In a recent case I was able to profile a writer as probably com-
ing from a Central European background, since his English misspelling shared
many characteristics with cognate words in Central European languages, but other
misspellings also indicated that the man spoke English with an Australian accent.
This narrowed the likely range of writers considerably.
A related area is that of trade names. Here the linguist may be asked whether
there is a likely confusion between two trade names, for example I was asked to
decide whether there was a possibility that two drugs, ‘Alkeran’ and ‘Arclan’ might
be confused in Australian English (particularly if ‘Alkeran’ were pronounced
beginning with a long ‘a’ [a:]). My conclusion, based in part on evidence of proc-
esses such as metathesis, and exchanges of [r] and [l] was that it was unlikely but
possible. Since such a confusion could be life threatening, this issue was important.
Similarly, Oyanadel and Samaniego (1999) were able to determine that the second
part of a trade name for baby cream ‘Fasaglos’ had been derived from an estab-
lished brand ‘Hipoglos’, by studying the morpheme -glos in Spanish. For a fuller
introduction to the trade name issue, see Butters (2008); Shuy (2002) provides an
extensive grounding.
In the area of vocabulary and grammar there are two main approaches used
in identification or profiling. The first is essentially probabilistic analysis, usually
performed by computer programs. There is a widespread belief, based in part on
literary studies, that there are certain grammar features and vocabulary choices
178 John Gibbons
that are used more by one person than another. It is important to note that this
works only when register variables such as topic, formality and genre type are held
constant, since these features also have a strong impact on both grammatical struc-
ture and vocabulary choice. Some supporters of such methods nowadays caution
against excessively strong statements based on them – see Grant (2008).
The second type of analysis is based on any peculiarities in grammatical structure
or vocabulary. Sometimes these are non-standard usages, and they may come from
a limited proficiency in either the register or the language that the person is using.
Coulthard (1997) and elsewhere has presented evidence that when police fabricate
evidence, they sometimes slip into police jargon and the hyper-elaboration dis-
cussed earlier. This phenomenon can be detected and revealed by a linguist. I have
testified that a transcript was not a faithful record of a second language speaker
because the transcript contained a range of tenses that he had not mastered.
Eagleson (1994) shows how a range of linguistic features, including spelling,
syntax, morphology and punctuation provided evidence concerning the authorship
of a letter which purported to be a suicide note. Police believed the letter had been
written by the woman’s husband, who was suspected of murdering her. Eagleson
compared samples of the husband’s and the wife’s writing, and was able to show a
range of features that were found in the man’s writing and the disputed letter, but
not in the woman’s writing, particularly ‘assult’ (for assault), ‘carring’ (carrying),
‘thier’ (their), and ‘treat’ (threat); the omission of the third person -s (e.g. ‘he
give’); the intrusive apostrophe; the omission of past tense -ed (e.g. ‘he never really
believe her’), and long poorly structured stretches of language with no punctua-
tion dividing them. The man changed his plea to guilty when confronted with this
evidence.
In the notorious Australian kidnapping case of Kerry Whelan, Robert Eagleson
and I were able to determine that the ransom letter, which masqueraded as com-
ing from an Asian gang, had probably been written by a native speaker of English
on the basis of the use of low-frequency elaborate vocabulary, and complex gram-
matical patterns. There were also indications from the patterning and format of
the letter that the writer may have had some experience in writing radio advertise-
ments (there were signs of intertextuality). This type of profiling, while it is not
conclusive in its identification, may avoid the expenditure of resources following
misleading indications.
Coulthard (1994) gave important evidence on cohesion phenomena to the
appeal of the Birmingham Six, which showed on the basis of the nature of the dis-
course that the police records of interview contained fabrication. For instance they
contained repeated reference to a ‘white plastic bag’ in that full form, rather than
beginning with the full form, and then using only ‘bag’ thereafter, which would be
normal in spoken discourse. This hyper elaboration is typical of legal language,
rather than everyday speech. Another feature was that the man consistently
referred to his friends by their first name, or their first name plus surname, while
in the contested samples, they were referred to by surname only. Coulthard also
examined a range of other features. The Birmingham Six were subsequently
released and paid compensation. The best evaluation of such evidence comes
when the person identified on linguistic grounds later confesses to producing the
language – see for example Eagleson (1994). There is a more complete discussion
of linguistic evidence in Coulthard and Johnson (2008).
9.7 Conclusions
Language and the law (sometimes also known as Forensic Linguistics) is an impor-
tant and fast developing area of real-world applied linguistic concern. All the issues
discussed here are of major significance to those involved, whether it is people
who cannot understand the legislation impacting on their lives, witnesses whose
testimony is distorted by linguistic pressure tactics, minorities whose language
cannot be used or who are subjected to group vilification or the guilty or innocent
convicted by language evidence. All these areas are open to examination and
action by applied linguists.
9.8 Prospects
This survey has concentrated mostly on the Linguistics of Forensic Linguistics.

However, books such as Rock (2007) hold out the possibility of an orientation
more rooted in the analysis of social process. She also shows the possibility of
further input from linguists in the area of police procedures. As corpora of legal
language and legal interaction develop, corpus based linguistic approaches such
as that of Heffer (2006) may provide new insights. As for Linguistic Evidence, the
possibility is emerging of a more coherent, scientific and conclusive approach – see
Coulthard and Johnson (2008). On the negative side, it is sadly the case that the
legal system continues to reject the long-term criticism from linguists of unneces-
sarily coercive courtroom questioning, particularly where vulnerable witnesses are
involved (see Eades, 2000 and 2008).
References
Atkinson, J. M., and Drew, P. (1979), Order in Court: The Organisation of Verbal Interaction
in Judicial Setting. London: Macmillan.
Bennett, W. L. and Feldman, M. S. (1981), Reconstructing Reality in the Courtroom.
London: Tavistock Publications.
Berk-Seligson, S. (1990), The Bilingual Courtroom Court Interpreters in the Judicial Process.
Chicago: The University of Chicago Press.
Bhatia, V. (1994), ‘Cognitive structuring in legislative provisions’, in J. Gibbons (ed.),
Language and the Law. London: Longman, pp. 136–155.
Bhatia, V. K. (1993), Analysing Genre: Language Use in Professional Settings. Harlow:
Longman.
Bowe, H. and Storey, K. (1995), ‘Linguistic analysis as evidence of speaker identifica-
tion: Demand and response’, in D. Eades (ed.), Language in Evidence Issues Confront-
ing Aboriginal and Multicultural Australia. Sydney, NSW: University of New South
Wales Press, pp. 187–200.
180 John Gibbons
Butters, R. (2008), ‘Trademarks and Other Proprietary Terms’, in M. T. Turell and J.

Gibbons (eds.), Dimensions of Forensic Linguistics. Amsterdam/Philadelphia: John
Benjamins (AILA series) pp. 231–247.
Carroll, J. (1995), The Use of Interpreters in Court. Forensic Linguistics, 2, (1), 65–73.
Cotterill, J. (2003), Language and Power in Court: A Linguistic Analysis of the O.J. Simpson
Trial. Basingstoke and New York: Palgrave Macmillan.
Coulthard, M. (1994), Powerful evidence for the defence: An exercise in forensic
discourse analysis, in J. Gibbons (ed.), Language and the Law. Harlow: Longman,
pp. 414–427.
Coulthard, M. (1997), ‘A failed appeal’, Forensic Linguistics, 4, (2), 287–302.
Coulthard, M. and A. Johnson (2008), An Introduction to Forensic Linguistics Language in
Evidence. London: Routledge.
Crystal, D. and Davy, D. (1969), Investigating English Style. London: Longman.
Danet, B., and Bogoch, B. (1994), ‘Orality, literacy, and performativity in Anglo-Saxon
wills’, in J. Gibbons (ed.), Language and the Law. London: Longman, pp. 100–135.
Danet, B., Hoffman, K. B., Kermish, N. K., Rafn, H. J., and Stayman, D. G. (1980), ‘An
ethnography of questioning’, in R. Shuy and A. Shnukal (eds.), Language Use and the
Uses of Language: Papers from the Fifth NWAV. Washington, DC: Georgetown University
Press, pp. 222–234.
Davies, E. C. (2004), ‘Register distinctions and measures of complexity in the language
of legal contracts’, in J. Gibbons, V. Prakasam, K. V. Tirumalesh and H. Nagarajan
(eds), Language in the Law. Delhi: Longman Orient, pp. 82–99.
Diamond, S. S. and Levi, J. N. (1996), ‘I’mproving decisions on death by revising and
testing jury instructions’, Judicature, 79, (5), 224–232.
Eades, D. (1994), ‘A case of communicative clash: Aboriginal English and the legal sys-
tem’, in J. Gibbons (ed.), Language and the Law. London: Longman, pp. 234–264.
Eades, D. (1995), ‘Language and the law: White Australia vs Nancy’, in M. Walsh and
C. Yallop (eds), Language and Culture in Aboriginal Australia. Canberra: Aboriginal
Studies Press, pp. 181–190.
Eades, D. (2000), ‘I don’t think it’s an answer to the question: Silencing Aboriginal
witnesses in court’, Language in Society, 29, 161–195.
Eades, D. (2008), ‘Language and disadvantage before the law’, in M. T. Turell and J.
Gibbons (eds), Dimensions of Forensic Linguistics. Amsterdam/Philadelphia: John
Benjamins (AILA series) pp. 179–195.
Eagleson, R. (1994), ‘Forensic analysis of personal written texts: A case study’, in
J. Gibbons (ed.), Language and the Law. Harlow: Longman, pp. 363–373.
Felker, D. B., Pickering, F., Charrow, V. R., Holland, V. M. and Redish, J. C. (1981),
Guidelines for Document Designers. Washington DC: American Institutes for Research.
Freedman, M. H. and Freedman, E. M. (eds) (1995), Group Defamation and Freedom
of Speech the Relationship between Language and Violence. Westport, CT: Greenwood Press.
Gibbons, J. (2001), ‘Revising the language of New South Wales police procedures:
Applied linguistics in action’, Applied Linguistics, 22, (4), 439–469.
González, R. D. and Melis, I. (eds) (2001), Critical Perspectives on the Official English Movement,
Vol. 2 History, Theory and Policy. Champaign-Urbana, IL: NCTE/Laurence Erlbaum.
Grant, T. (2008), ‘Approaching questions in forensic authorship analysis’, in M. T. Turell
and J. Gibbons (eds), Dimensions of Forensic Linguistics. Amsterdam/Philadelphia:
John Benjamins (AILA series) pp. 215–229.
182 John Gibbons
Papke, D. R. (ed.) (1991), Narrative and Legal Discourse a Reader in Storytelling and the
Law. Liverpool: Deborah Charles Publications.
Phillips, S. (1987), ‘On the use of Wh questions in American courtroom discourse:
A study of the relation between language form and language function’, in L. Kedar
(ed.), Power through Discourse. Norwood, NJ: Ablex, pp. 83–111.
Rock, F. (2007), Communicating Rights the Language of Arrest and Detention. Basingstoke
and New York: Palgrave Macmillan.
Shuy, R. W. (1993), Language Crimes the Use and Abuse of Language Evidence in the Court-
room. Oxford: Blackwell.
Solan, L. M. (1995), ‘Judicial decisions and linguistic analysis: Is there a linguist in the
court?’ Washington University Law Journal, 73, (3), 1069–1083.
Steinberg, E. T. (ed.) (1991), Plain Language – Principles and Practice. Detroit, MI: Wayne
State University Press.
Stygall, G. (1994), Trial Language: Differential Discourse Processing and Discursive Formation.
Amsterdam and Philadelphia: John Benjamins.
Taylor, B. (1995), ‘Offensive language: A linguistic and sociolinguistic perspective’, in
D. Eades (ed.), Language in Evidence: Issues Confronting Aboriginal and Multicultural
Australia. Sydney: UNSW Press, pp. 219–258.
van Dijk, T. A. (1987), Communicating Racism: Ethnic Prejudice in Thought and Talk.
Newbury Park, CA: Sage.
van Dijk, T. A. (1993), Elite Discourse and Racism. Newbury Park, CA: Sage.
Vasilachis de Gialdino, I. (1997), Discurso Político y Prensa Escrita. Barcelona: Editorial
Gedisa.
Vlachopoulos, S. (2004), ‘Translating the untranslatable? The impact of cultural con-
straints on the translation of legal texts’, in J. Gibbons, V. Prakasam, K. V. Tirumalesh
and H. Nagarajan (eds), Language in the Law. Delhi: Orient Longman, pp. 100–115.
Wallace, C. (1990), Reading. Oxford: Oxford University Press.
Walsh, B. (1995), ‘Offensive language: A legal perspective’, in D. Eades (ed.), Language
in Evidence Issues Confronting Aboriginal and Multultural Australia. Sydney: UNSW
Press, pp. 203–218.
Walsh, M. (1994), ‘Interactional styles in the courtroom’, in J. Gibbons (ed.), Language
and the Law. Harlow: Longman, pp. 217–233.
Weaver, C. (1988), Reading Process and Practice: From Socio-psycholinguistics to Whole
Language. Portsmouth: Heinemann Educational.
Further Reading
General
Gibbons, J. (ed.) (1994), Language and the Law. London: Longman.

Gibbons, J. (2003), Forensic Linguistics an Introduction to Language in the Justice System.
Oxford: Basil Blackwell.
Levi, J. N. and Walker, A. G. (eds) (1990), Language in the Judicial Process. New York:
Plenum.
The International Journal of Speech Language and the Law.
Legal Language
Tiersma, P. M. (1999), Legal Language. Chicago: The University of Chicago Press.
Courtroom Interaction
Cotterill, J. (2003), Language and Power in Court: A Linguistic Analysis of the O.J. Simpson
Trial. Basingstoke and New York: Palgrave Macmillan.
Heffer, C. (2006), The Language of Jury Trial: A Corpus-Aided Linguistic Analysis of
Legal-Lay Discourse. Basingstoke and New York: Palgrave Macmillan.
Language and Disadvantage before the Law
Eades, D., (ed.) (1995), Language in Evidence: Linguistic and Legal Perspectives in Multicul-
tural Australia. Sydney: University of New South Wales Press.
Vilification
Freedman, M. H. and Freedman, E. M. (eds) (1995), Group Defamation and Freedom of

Speech the Relationship between Language and Violence. Westport, CT: Greenwood
Press.
Language Crimes
Shuy, R. (1993), Language Crimes the Use and Abuse of Language Evidence in the Courtroom.
Oxford: Blackwell.
Linguistic Evidence
Hollien, H. (2001), Forensic Voice Identification. New York: Academic Press.

Coulthard, M. and Johnson, A. (2008), An Introduction to Forensic Linguistics Language in
Evidence. London: Routledge.
CHAPTER
10
Neurolinguistics and
the Non-monolingual Brain
Marjorie Lorch
10.1 Introduction
Neurolinguistics is an area of research which came into its own towards the end of
the twentieth century. Researchers working in this field try to understand the way
language is organized in the brain, taking the performance of people with neuro-
logical impairments as the main source of evidence (Whitaker and Whitaker,
1976–1979; Goodglass, 1993; Caplan, 1987; Ahlsen, 2006). The conceptual basis of
this field was established over 100 years earlier with the clinical and theoretical
work on localization of function in the cortex by Broca, Wernicke, Liepmann, etc.
(Eling, 1994).
A primary source for research from the 1960s onwards was the Veterans
Administration Hospitals in the United States. This clinical setting provided a large
group of typically white male English-speaking neurological patients who had suf-
fered strokes in their 60s. In addition, they had long-stay chronic care arrange-
ment in medical centres with interdisciplinary teams of health care and research
professionals.
This particular medical context created the opportunity to make detailed obser-
vations of spared and impaired language, speech and voice functions in people
over a long period of time and correlate this with detailed neurological and psy-
chological assessments. For the next 30 years, groups of stroke patients were
studied by clinical researchers who had the benefits of insights from Chomsky’s
theoretical developments in Linguistics (e.g. Goodglass and Blumstein, 1973;
Caplan and Hildebrandt, 1988; Kean, 1985; Grodzinsky, 1990; Bastiaanse and
Grodzinsky, 2000). Much was gained through this confluence of opportunities.
Bedside examinations, which were the rule-of-thumb approach used by clinical
neurologists, have given way to formal assessment techniques developed by psy-
chologists and speech and language pathologists. These new instruments had the
virtue of being standardized with normative values based on observations of large
populations samples (e.g. Boston Diagnostic Aphasia Examination, Goodglass and
Kaplan, 1976, Minnesota Test for Differential Diagnosis of Aphasia (Schuell, 1965),
Porch Index of Communicative Ability, (Porch, 1971), Western Aphasia Battery,
(Kertesz, 1982), etc).
Neurolinguistics and the Non-monolingual Brain 185
Correlations between clinical descriptions and neurological pathology could be

made through innovations in neuroimaging techniques which were also develop-
ing in tandem with other clinical assessments. From the advent of radioactive scan-
ning (e.g. Benson, 1967), to CAT scans (e.g. Naeser and Hayward, 1978) BEAM
(e.g. Duffy, 1985), PET and MRI (e.g. Rapcsak et al., 1990) and more recently MEG
(e.g. Breier et al., 2006), technological innovations in computing and biophysics
have allowed for dynamic clinico-pathological investigations in living patients
whereas previously, lesion confirmation could only be gained by autopsy.
At the turn of the twentieth century, a new research landscape is beginning to
take shape. In the decades of research following the breakthroughs in the 1960s,
our understanding of language functioning was derived from the profile observed
in a narrowly defined homogenous group, primarily consisting of right-handed
English speaking 60-year-old men with high school education who had suffered
a single cardiovascular accident (CVA). These research observations had built up
a ‘normal’ prototype which informed our assumptions and hypotheses about how
language was organized in the brain. The current picture reflects both changes in
the opportunities for and access to neurolinguistic evidence, and the diffusion of
North American research training throughout the world. New research is now
focusing on a range of variables which have been identified as crucial to our under-
standing of how human language is organized in the brain.
The objective of this chapter is to provide a synthesis of neurolinguistic research,
to provide a state-of-the-art review and to give some examples in the way in which
research may be fruitfully directed in future. Consideration will be given to the
development of our understanding of how language is organized in the brain by
refinements in both the way research questions are framed and the manner in
which answers are sought. Any domain of active research investigation undergoes
refinement over the years, from observation to classification, defining categorical
distinctions and generating hypotheses. With the move to experimental research,
refinement is achieved through the identification of relevant sources of variation
in populations and task properties which affect the nature of observations. Under-
standing of the sources of heterogeneity in human beings with respect to the
organization in the brain has been served by the increasing number of factors
which have been demonstrated to correlate with patterns of function.
10.2 Key Issues Emerging from Existing Research
10.2.1 The Importance of Identifying Subject Variables
In the 1950s, a book was published which constituted a major attempt to character-
ize how language impairments were produced by brain pathology. The series of 155
aphasic cases was studied with respect to diagnosis and prognosis. It was collected
by Hildreth Schuell and her colleagues at the Minneapolis Veterans Hospital
(1964) and can be seen as a major benchmark of modern aphasia research (see
Weisenberg and McBride, 1935 for an earlier attempt). At this time, the subject
variables were defined as age, education level, occupation, handedness, etiology
186 Marjorie Lorch
(i.e. cause of illness) and time post onset of illness, with hearing disorders, devel-
opmental speech difficulties and psychiatric problems excluded. The neuropatho-
logical evidence was supplied by electroencephalographic recordings (EEG).
The interpretative meaning of the first subject variable – that of age – has grown
in significance recently. This arises from more sophisticated understanding of
maturational issues with reference to the development of the language faculty.
Advances have been made both at the neurological level, with increasing knowl-
edge of processes of cell migration and death, myelination patterns, etc. (e.g.
Geschwind and Galaburda, 1984), in more complex theories about the effects of
experiential stimulation from the environment (Karmiloff-Smith, 1992) and the
interaction of factors under genetic control (Vargha-Khadem et al., 2005).
Throughout much of the nineteenth century, age was not recognized as a signif-
icant issue with regard to behaviour and the brain. Towards 1900, there was increas-
ing awareness that patients had to be differentiated into at least two groups, that of
children and adults. Work in developmental psychology and clinical neurology
provided refinements which made meaningful neurolinguistic distinctions between
infants, school children, post-puberty adults and the elderly (Hellal and Lorch,
2009). Interest in how maturation of the nervous system responded to injury at dif-
ferent periods of development of the language faculty grew (Lenneberg, 1967).
The acquisition of language is now seen as a dynamic process that unfolds over
a long period of time. For a detailed understanding of how language capacity
develops in an individual, maturational issues now appear to be central to an ade-
quate account. New understanding of genetics on the one hand, and increasing
research on the healthy aging brain with the recent extension in lifespan expect-
ancy on the other, has also given rise to a new appreciation that with respect to how
language is organized in the brain, age matters.
While, as discussed above, the foundations of the modern science of neurolin-
guistics was based on adults with acquired disorders, new research on Specific
Language Impairment (Gopnik and Crago, 1991) and other developmental disor-
ders which affect language such as Williams Syndrome (Bellugi et al., 1997), Down
Syndrome and Autism (Tager-Flusberg, 1994) provide another source of impor-
tant evidence (Jenkins, 2000).
Although some of this research is carried out with groups of individuals defined
by their disorder, other research questions arise in the context of individual cases
which display unusual patterns. For example, Atkin and Lorch (2006) studied a
4-year-old boy with a highly developed reading ability (hyperlexia) in the context
of a profound impairment in cognitive development (Autistic Spectrum) and the
absence of spoken language. He only produced spoken language in response to
written stimuli, but no evidence of language comprehension could be demon-
strated through writing, speech, pictures or gestures. However, the boy produced
correct heterographic homophones (here/hear) in context. In addition, he made
semantic paraphrases while reading which maintained syntactic consistency across
words. These two findings imply a level of linguistic development far beyond that
which would be predicted by the child’s mental age of one and a half years. The
observations in such a case indicate the possibility of an atypical route to language
acquisition and the development of literacy for which existing cognitive accounts
are inadequate. In order to account for this child’s behaviour, new developments
in theoretical models may be called for. Unusual developmental trajectories such
as this will be an important and growing source of neurolinguistic evidence in
future.
The identification and study of such individuals is serendipitous. It relies on the
presence of neurolinguistically sophisticated researchers to be available to schools
and clinics. Individual researchers and teams working within education and health
care can provide important observations for theory development, but only if they
are sufficiently grounded in neurolinguistics to identify atypically behavioural phe-
notypes or symptom complexes.
10.3 The Variable of ‘Time Post-Onset of Illness’
The focus in traditional neurolinguistic research has been on adult acquired neuro-
genic disorders of language. The majority of aphasic subjects studied had suffered
cerebrovascular accident after acquiring language normally. In the twentieth cen-
tury, this was a major source of illness in Western populations. (In Schuell’s case
series the frequency was 83 per cent, Schuell et al., 1964.) Properties of this neu-
rovascular insult define the quality of behavioural impairments observed clinically.
That is, by their nature, strokes are singular, acute events which have a pattern of
recovery determined by how the nervous system responds to the changes in blood
supply on the day, over the next few days, in the first two months and chronically.
The understanding of how recovery unfolds over time in response to a cardiovas-
cular event is an active area of present clinical research (Hillis et al., 2001). The
variable ‘time post onset of illness’ for stroke patients can now be interpreted in
a more meaningful way.
Alternatively, in other etiologies causing chronic, relapsing or progressive illness,
the notion of ‘time post onset of illness’ takes on a completely different status. These
etiological properties will determine some of the clinical picture in aphasia. If, for
example, a person has a language disorder as a result of a tumour, which is slow
growing and space occupying, the picture will necessarily be quite different. Our
clinico-pathological correlation approach to language disorders and brain function
has been based on the properties of impairment observed in stroke. This approach
capitalized on the singular and acute aspects of CVAs. Patients suffering from other
illnesses are now beginning to provide a more elaborated picture of the organiza-
tion of language through these distinct patterns of impairment. Alzheimer’s disease
(Nicholas et al., 1985), Primary progressive dementia (Hodges et al., 1992), Multi-
ple Sclerosis (Devere et al., 2000), Parkinson’s disease (McNamara and Durso,
2003), Spasmodic Dysphonia (Whurr et al., 1993), etc., all provide sources of evi-
dence for how thought, language, speech and voice are produced in the brain.
The reduction in the frequency and severity of strokes has been achieved in
Western populations through a series of public health advances. At the same time,
progressive diseases in the elderly are being studied more intensively as the general
188 Marjorie Lorch
population enjoys increased longevity. In future, the issue of how different disease
processes affect neurological system functions will ultimately prove to be a signifi-
cant issue.
Cause of illness (i.e. etiology) has typically been a variable which is reported
as a subject variable but not often incorporated into the interpretation of data.
More sophisticated understanding of how functions are impaired and recover
from various neurological pathologies will provide better accounts of their neuro-
linguistic performance.
Other variables of current neurolinguistic interest are gender and handedness
differences. (Note that gender was not included in Schuell et al.’s 1964 list perhaps
due to the predominance of male Veterans subjects.) Studies of women suggest
that mental faculties and brain organization may vary, although there is great
debate as to whether this may be accounted for by learned social roles, genetic
factors or some interaction of the two (Kimura, 1999). Similarly, some left-handed
individuals appear to have different patterns of cognitive organization. Studies of
hemispheric specialization and lateralization of function were prevalent in neuro-
linguistic research in the latter part of the twentieth century (Benson and Zaidel,
1985). The strong predictive relationship between left hemisphere pathology and
aphasic speech has been documented for over hundred years (Schwartz, 1894).
However, a more complex view has developed with the study of extralinguistic impair-
ments in communication in people suffering right hemisphere damage (Code,
1987). This line of research has been assisted by the growing fields of pragmatics
and emotional processing in relation to the social functions of speech (Paradis,
1998). At the same time, recent imaging studies indicate bilateral participation in
the language processing of healthy individuals and right hemisphere involvement
in recovery of language functions in aphasic individuals (Cappa and Vallar, 1992).
The development of different patterns of brain organization for cognitive
functions including language has been emerging on another front. Prescient
work initiated by Norman Geschwind (d. 1985) and colleagues on handedness, the
development of particular cognitive abilities, epigenetic factors in foetal brain
development, and their consequences in later life such as dyslexia (Geschwind and
Galaburda, 1985) have recently begun to be pursued with renewed vigour. For
example, arising out of his work in Autism, Baron-Cohen (2002) has detailed a
continuum of cognitive behavioural patterns linked with genetic factors for the
normal population. The notion of multiple intelligences, developed over the past
two decades by Howard Gardner (1999), has also fostered a view of the language
faculty as one part of a mosaic of differentially developing mental capacities driven
by genetic and environmental influences.
10.4 The Language Variable
Although the research landscape reviewed above showed a positive trend in poten-
tial developments for understanding in neurolinguistics, one crucial variable has yet
to be widely recognized, i.e. the language or languages a person speaks. In Schuell
et al.’s case series (1964) no patient information was recorded on this variable.
It was either assumed that an individual was an English-speaking monolingual, or
that the knowledge of other languages was irrelevant to their language disorder in
this clinical setting. All assessment and treatment would take place in English. No
notice was taken if a person had grown up speaking another language, or in fact
predominantly spoke something other than English outside of the hospital setting.
Although bilingual aphasia cases were reported in the literature (first in 1895 by
Pitres), throughout much of the twentieth century, there was little theoretically
driven interest in how being multilingual might affect how their languages were
represented in the brain (Lorch, 2008). In addition, there was a lack of research
into how damage to the brain might affect languages with grammatical properties
which were distinct from those exhibited by English.
In the 1980s, a number of researchers with training in both linguistics and
neuropsychology began to realize that aphasia research which focused exclusively
on English-speaking monolingual individuals could not afford an adequate descrip-
tion of the human language faculty (Chiarello et al., 1982, Paradis et al., 1985).
Cross-linguistic aphasia projects were carried out on monolingual individuals in
different countries by researchers with training in the North American tradition
(Bates and Wulfeck, 1989; Menn and Obler, 1990). This ensured consistency
in methodology and clinical practice. Interest in questions regarding language
universals, which was growing in the domain of theoretical linguistics, helped
to motivate this research (Comrie, 1981). This research effort was also aided by
better grammatical descriptions of languages other than English that were emerg-
ing at the same time in the Principles and Parameters framework of generative
grammar (Chomsky, 1981). In addition, the increased professionalization and
academic training of speech therapists worldwide at the end of the twentieth cen-
tury has contributed to more published research on language disorders in people
speaking languages other than English. In parallel, there has been a flowering of
research into the neurolinguistic properties of visual-gestural languages with the
growth in social prominence of the Deaf communities (Poizner et al., 1987). In
addition, increasing multilingualism and the socio-political drivers to provide
health care and education for people who speak languages other than that of the
dominant population has changed the potential for neurolinguistic research.
Modern bilingual aphasia research was instigated by a number of researchers
in the 1970s such as Paradis (1977) studying the bilingual culture in Montreal
Canada, and Albert and Obler (1978) multilingual researchers in Boston. Through-
out the next two decades, there was a great deal of research activity investigating
bilingual aphasia. This topic was regarded as an important source of evidence for
understanding the general relationship between brain organization and language
processing.
The value of bilingual aphasia as evidence for neurolinguistic theory is due to
the huge variability in the picture of clinical symptoms and patterns of recovery
that have been documented. One would expect that if all a multilingual speakers’
languages were processed in the same way, brain damage which led to language
impairment would affect them equally. Surprisingly, there appears to be a substantial
190 Marjorie Lorch
minority of individuals who do not show this pattern of language impairment.

There may be different types of aphasic symptoms in the different languages, dif-
ferent levels of severity of aphasic symptoms in the different languages, and/or
different rates of recovery of aphasic symptoms in the different languages. It
appears that languages in bilinguals may be psychophysiologically distinct, with
languages being selectively impaired and preserved.
Although many aspects of aphasia appear to be fairly consistent and predictable
(such as the strong association between anterior lesions with production difficul-
ties and posterior lesions with comprehension problems), the patterns of impair-
ment documented in non-monolinguals are difficult to interpret. This is unexpected
and surprising. It suggests that people who learn to use more than one language
do so in many different ways, and that the mental representation of those lan-
guages can have a variety of different forms and functions.
A number of different factors have been suggested to explain the patterns
seen in impairment and recovery of bilingual aphasics: (1) a patient’s native lan-
guage would recover before languages learned later (Ribot, 1881), (2) ‘. . . [the
language] the most familiar to the patient (usually, but not always, the mother
tongue) . . . reappears first because it is the one that uses the most solidly fixed
associations’ (Pitres, 1895: 47), (3) the order of recovery seen in bilingual aphasia
based on degree of automatization (Pick, 1921/1973); (4) recovery of languages
depended on affective and emotional factors (Minkowski, 1927). More recently,
other social aspects of communication have also been considered to be of rele-
vance to the picture of language impairment in non-monolingual speakers: (1) the
language of the patient’s present environment, (2) the individual’s communica-
tive needs, (3) the individual’s literacy attainment and practice, (4) the domain of
communication (i.e. related to work or personal topics, etc.), (5) the language of
the clinical environment and of therapy delivery, (6) the degree of structural dif-
ference between the grammars of the languages spoken etc.
Paradis (1977) devised a classification of bilingual aphasia recovery patterns:
z Parallel – similarly impaired and recover at the same rate

z Differential impairment and recovery – of a different degree in the different
languages relative to the pre-morbid mastery
z Successive – one language does not begin to reappear before another is maxi-
mally recovered
z Antagonistic – one language regresses as another progresses
z Selective – recovery occurs in one language but not in another
z Blended or mixed – inappropriate mixing of two or more languages (not equiv-
alent to normal code switching)
z Alternating antagonism – for alternating periods of time the patient has access
to only one of their languages.
All logically possible patterns of impairment and recovery have been documented.
Twenty-four years later, Paradis (2001) carried out a review of all cases of bilingual
aphasia published between 1985 and 2000. In the 132 bilingual aphasic cases that
had been reported, there were 81 with parallel recovery, and 24 with differential
recovery. Of those showing differential recovery of their languages, 12 individuals
displayed language mixing, 9 showed selective impairment in one language with
respect to the other(s) and 6 showed successive recovery of their languages.
After reviewing all possible variables which might explain this pattern of
results, Paradis concluded that none of the factors identified could account for the
observations: ‘Neither primacy, automaticity, habit strength, stimulation pre- or
post-onset, appropriateness, need, affectivity, severity of aphasia, type of bilingual-
ism, type of aphasia nor structural distance between the languages could account
for all the non-parallel recovery patterns observed’ (Paradis, 2001: 77).
Current approaches to understanding non-monolingual language functions
draw on psycholinguistics processing models which employ metaphorical compu-
tational instantiations of activation and inhibition of nodes in networks (Green,
1986, 1993, 1998). In addition, there have been attempts to understand the variety
of ways in which a person might become multilingual within a framework drawing
on current models of working memory. Paradis’ model is based on ideas about
short versus long-term memory, and implicit versus explicit memory systems.
Paradis has put forward the hypothesis that that the mother tongue and the sec-
ond language may be subserved by different memory subsystems. The acquisition
of the mother tongue is thought to rely more on implicit procedural memory,
while second language learning after the age of seven in a classroom will rely more
on explicit declarative memory. These two types of memory are known to be neuro-
anatomically distinct.
It has become increasingly clear that limitations in our understanding of how
non-monolingual speakers process multiple languages are to some extent based
on the difficulty in identifying and classifying the neurolinguistically relevant
features of individuals’ life history in the learning and use of multiple languages.
That is, in order for this strand of neurolinguistic research to make any progress,
new subject and task variables must be identified.
Bilingualism research in sociolinguistics has suggested a number of variables
which appear to impact on the way speakers use different languages in different
settings, e.g. at home, work, social and community (Grosjean, 1982). In addition,
the manner of language learning is an area of active research in applied linguistics.
Such issues which are currently being examined explore a variety of dimensions
including: how an infant is exposed to languages at home, study and instruction
methods in school, the role of the individual in relation to properties of the social
context such as identity and attitudes, group status, group size, etc. Levels and
domains of proficiency and attainment and aspects of literacy are also relevant
factors. All of these factors shape individual language histories which will be
reflected in different neurolinguistic instantiations. In addition, differential use
of languages on a day-to-day basis may also lead to significant effects. Regular
use of codeswitching in integrated bilingual communities may represent a type of
neurolinguistic functioning quite distinct from a person who used one language
as an infant and another for the remainder of their adult life as an immigrant to
a new language community.
192 Marjorie Lorch
Curiously, although multilingual communities are increasing in major urban

areas of Western countries, this has had little impact on clinical practice. Paradis
noted that in the United States,
whereas 25 years ago, the fact that a patient spoke a language other than that of
the hospital environment was rarely recorded in that patient’s file, today at least
one course in bilingualism is required in language pathology training programs,
and patients are increasingly assessed in more than one language. (Paradis,
2000: 179)
This seems to be a minimal change in practice in response to such a large social

change.
Recent appraisals of the field of bilingual aphasia research have been discoura-
ging. In a review of the current state of neurolinguistics, Ahlsen states ‘Neurolin-
guistic aspects of bi or multilingualism have only been studied extensively by a few
researchers’ (2007: 127). In another recent overview of research findings in this
area, Fabbro (2002: 202–203) asserted that, ‘It can thus be concluded that so far
empirical studies have not provided tenable explanations for the presence of par-
allel recovery in some bilingual aphasic patients and of differential recovery in
others.’
10.5 Imaging the Brain
Although research into non-monolingual aphasia has yet to realize a new way for-
ward, there is great interest in bilingualism with respect to processing research in
healthy individuals. The advent of imaging techniques such as functional magnetic
resonance imaging (fMRI) has afforded the opportunity to investigate the localiza-
tion of languages in the healthy bilingual brain. For example, Mechelli and col-
leagues (2004) claim that the grey matter in Broca’s area increases in bilinguals
relative to monolinguals, especially in those who learned a second language early
in life. ‘It reinforces the idea that it is better to learn early rather than late because
the brain is more capable of adjusting or accommodating new languages by chang-
ing structurally. . .’(Mechelli et al., 2004).
The cerebral localization of multiple languages by using imaging techniques in
healthy individuals appears to be a topic of active research. To date, over 40 imag-
ing studies with healthy non-monolingual speakers have been published. A recent
meta-analysis of this research led Hull to state that
Unfortunately . . . very few of these studies have been designed in a way to allow
comparisons of bilinguals with monolinguals, or of bilinguals with other bilin-
guals differing in age of onset of language exposure, thereby making this source
of evidence not very informative about individual differences in brain organiza-
tion related to language experience. (Hull, 2003: 12)
This underscores again the importance of determining the theoretical status of

such subject variables with precision.
10.6 A Research Strategy of Converging Evidence
The need for renewing our attempts to identify sources of converging evidence
should shape the future research landscape into neurolinguistics. Our accounts
of clinical observations from people with language impairments must be compati-
ble with understanding that is gained through (1) developments in anatomy and
physiology to characterize the function of the nervous system, (2) social and cogni-
tive psychology which models behaviours of healthy individuals and (3) theoretical
linguistics which provides accounts of grammar. Insights from these subjects should
inform the questions we ask of aphasic subjects. It is often the case that patterns of
spared and impaired performance in particular task domains have gone undocu-
mented because investigators had not thought to examine them. Stimulus materi-
als, task design and assessment materials must reflect the current state of theory.
For example, current investigations which employ tasks comprised of single
words are common in imaging studies of language processing due to technological
constraints on scanning. However, it is clear that single words have particular prop-
erties which means that they may not represent the functioning of language sys-
tems per se. An argument put forward on this point by Paradis (2004) is a good
example of the types of converging evidence which should be used to inform our
investigative strategies.
. . . single words are the least likely candidates for investigating ‘language’ repre-
sentation, given that what makes language most specific as a cognitive function,
i.e., the language system (phonology, morphology, syntax), is supported by
procedural memory, whereas isolated words are supported by declarative
memory and are hence less focalized in their cortical representation. (Paradis,
2004: 173)
In order to support his argument, Paradis lists a number of different sources of

evidence for backing up this claim: (1) the ease of word learning but not syntactic
acquisition by non-humans, (2) the limited acquisition of syntax in contrast to
lexical development as a consequence of deprivation in early childhood, (3) the
difficulty of learning new words but not procedures by anterograde amnesic
patients such as H.M., (4) the difficulties of word retrieval but not syntax in those
suffering from Alzheimer’s dementia, (5) loss of lexical but not syntactic perform-
ance in first language attrition, (6) the relative ease of vocabulary learning in late-
second-language students relative to difficulty with phonology, morphology and
syntax (Paradis, 2004). Words presented in isolation lose the linguistic properties
that only exist in particular constituent contexts. This notion of the unit of the
sentence as prime for any linguistic characterization has been accepted since
the middle of the last century (Chomsky, 1965). Researchers who use single word
stimuli must not believe that their results will have any direct bearing on explana-
tions regarding language processes per se.
Consider the currently widely used clinical task of ‘word fluency’. This requires
one to produce as many words as possible in a given amount of time with a particular
‘phonological’ (i.e. orthographic) or ‘semantic’ (i.e., world knowledge) property
194 Marjorie Lorch
such as beginning with the letter ‘f’ or animals. This is not a task which reveals
anything about the language faculty in and of itself. It is used as a clinical diagnos-
tic tool with patients suffering from a diverse group of disorders which involve
cognitive but not language impairments (e.g. Neurofibromatosis 1, Lorch et al.,
1999). Poor performance is interpreted with respect to attention and executive
control functions not linguistic processes.
In addition to crucial aspects of a research strategy pertaining to task selection,
issues regarding the status of individual differences must also be considered. The
tendency towards group studies with an emphasis on homogeneity of subjects
which prevailed in the mid-twentieth century has been steadily replaced by a
preference for single case studies. The type of strategy based on explanations of
behaviour in single cases has been the hallmark of the cognitive neuropsychology
research agenda (Shallice, 1988). It asserts the value of an account which is theo-
retically coherent for an individual pattern of spared and impaired behaviours in
acquired disorders. By generating subject variables which are theoretically derived,
more convergence of findings based on individual cases may be achieved.
10.7 Looking Back To Go Forward
Finally, current theoretical developments and research strategies may be enhanced

by employing an applied historical perspective. A historical analysis may reveal the
social and cultural context in which current theory and methodology was initially
developed. It may also lead to the development of a more comprehensive approach
for current research. How we perceive a problem and characterize variables will
directly affect our ability to find solutions. For example, bilingual aphasia appears
to be a crucial area of neurolinguistics in which to address questions about biobe-
havioural concomitants of language experience. However, as the review presented
above demonstrates, current researchers are at a loss to provide a coherent expla-
nation of the body of research findings which have been amassed over the past
three decades. Insights from a historical perspective may provide novel approaches
to such areas of research which are currently at an impasse.
A recent applied historical approach to the characterization and etiology of
Tourette’s syndrome is a ground-breaking piece of research. The development of
vocal tics and involuntary swearing does not currently have a theoretically moti-
vated explanation. Kushner (1999) went back to the original nineteenth century
clinical reports and found that a consistent association of rheumatic fever with
the subsequent development of tics and vocalizations was noted in this earlier
literature. On the strength of this observation, Kushner initiated lab research
which identified post-infectious streptococcus antibodies as one source of the neu-
ropathology leading to Tourette’s syndrome. Kushner states the rationale for an
applied historical approach as follows:
On the one hand, such investigations provide an illustration of how a historical

interrogation of syndrome construction can free medical researchers to pursue
alternative and novel approaches. On the other hand, they demonstrate how
historians can make unique contributions as collaborators in clinical care and
medical research . . . Historical investigations of syndrome construction can
elicit useful issues for the development of research hypotheses and clinical
diagnoses. In this way, applied historians of medicine can become important
partners in collaborative interdisciplinary medical research. (Kushner, 2003)
I have considered the problem of characterizing the patterns observed in bilin-

gual aphasia using an applied history approach. The aim is to reveal some new ave-
nues for research into language disorders in non-monolingual speakers. Paradis’
(1983) extensive review of the historical literature identified a paper by Pitres
(1895) as being the first documented case of bilingual aphasia. This is surprising,
since there were thousands of publications recording cases of acquired language
disorders in monolingual speakers after Trousseau coined the term aphasia in
1864. The question I raise is why no cases of bilingual aphasia were recorded
in that 30-year period of active research? In the archival study I recently carried
out (Lorch, 2006a; 2008), no records were found in the English medical literature
of cases identified as bilingual aphasia before Pitres (1895). What was revealed
through my investigation were detailed descriptions of cases where one language
was differentially affected subsequent to neurological illness in non-monolingual
speakers, but these patients were categorized and conceptualized completely dif-
ferently by nineteenth-century practitioners. These cases were not clinically classi-
fied as ‘loss of speech’ (the term for aphasia pre-1864), but rather as memory
disorders. It appears that in the nineteenth century, knowledge of second lan-
guages was considered to have a different psychological status than today. They
were assumed to be a reflection of a general intellectual achievement rather than
something pertaining to the language faculty. Learning a second language was
categorized as an academic endeavour akin to learning geography or science. This
nineteenth century characterization of the problem drew a clear distinction
between speech disorders in monolinguals as opposed to memory impairment in
bilinguals. This historical demarcation throws into relief some of our own current
conceptualizations of the difference between first language acquisition of a native
language and other types of learning. In order to develop our ideas about lan-
guage organization in non-monolinguals we may need to revise our assumptions
about the contribution of ‘learning’ and ‘memory’. With our growing under-
standing of the neurological underpinnings of such processes, more sophisticated
neurolinguistic models may emerge concerning how an individual becomes
non-monolingual.
The cultural learning of more than one language by an individual is generally
considered as exceptional in neurolinguistic terms from an Anglo-American
perspective. As highlighted in the review above, current formulations take the
monolingual individual as the neurolinguistic norm to be modelled. What is
also considered to be the neurolinguistic norm in this research domain is that of
literacy. The cultural learning of an orthographic representation of spoken lan-
guage has gained such primacy in our society that the possibility that the typical
196 Marjorie Lorch
neurolinguistic state of humans, from a more global, historical and anthropolo-

gical perspective, is more likely to be multilingual and illiterate. Nevertheless, a
large portion of modern research into language processing is focused on the neu-
rolinguistic substrates for reading and writing.
An applied history approach to the neurolinguistic substrates of literacy have also
revealed interesting shifts in our theoretical conceptualization of such processes.
My research into the nineteenth and twentieth century literature on the acquired
disorder of written language production reveals some interesting changes in the
cultural value placed on this cognitive ability and how cultural learning is thought
to have direct effects on the specialization of the brain (Lorch and Barrière, 2002,
2003). In the class of acquired disorders of cognitive functions, pure agraphia
represents something of an anomaly. It was first described in the modern literature
by Pitres in 1884 as a selective loss of the ability to produce written language in the
context of spared spoken language and reading ability. Over the next 125 years,
papers have been published documenting other cases of this rare clinical picture
while, at the same time, others have made statements in print denying its existence.
Questions have been raised about the robustness of observations. At the same time,
the possibility of its existence has been rejected a priori on theoretical grounds.
This raises issues regarding rareness of an observation which by its compelling
nature is viewed as requiring an explanatory account on the one hand, or as an
exception ruled as a methodologically flawed piece of outlier data (Barrière and
Lorch, 2004).
Following on from the analysis of pure agraphia, my historical investigations
probed the methodological status of clinical observation and the potency of another
emblematic case in the historical literature which became a source of evidence for
significant theoretical argumentation. Hellal and Lorch (2007) reviewed the case
of a child with an unusual pattern of acquired language impairment, recovery and
impairment, who was reported to have had homologous lesions in left and right
Broca’s area at his death. This case, published by Thomas Barlow in 1877, was cited
in the literature for over three decades as definitive evidence in support of a
number of theories regarding the role of the left hemisphere for language, and
the right hemisphere in the recovery of function.
We compared the record as published in the British Medical Journal with those
of the hospital case notes archived at the Great Ormond Street Hospital, London.
We found a number of fundamental discrepancies between the clinical case notes
and the journal article belying claims about the recovery of his language impair-
ment and the size and locations of the lesions. These were compounded by subse-
quent authors when citing this case. This dubious evidence was nevertheless used
by leading theoreticians to support arguments about language function, develop-
ment of dominance and recovery patterns in children. These notions were so per-
vasive in the late nineteenth and early twentieth century that this one, poorly
reported case was sufficient to persuade people of the validity of these theories.
Our review of this case also raises questions regarding observational and recording
practices which are relevant to today’s clinical researchers. Other issues which have
been illuminated by this type of applied historical neurolinguistics are the evolving
concepts of organic versus functional disorders and methodological issues (Lorch,

2006b), the changing status of behavioural evidence for diagnosis (Lorch, 2006c)
and how subject variables get identified (Hellal and Lorch, 2009).
This chapter has reviewed the state of the art research which has given rise to
universal biolinguistic accounts for how language is organized in the brain. How-
ever, I point out that these accounts are predicated on research observations
primarily drawn from monolingual literate individuals. Recent research has begun
to actively investigate the organization of language in non-monolingual speakers
using both clinico-pathological lesion studies and imaging techniques in neuro-
logically impaired and healthy adults and children. There has been only limited
progress in the interpretation of these findings towards the development of
an account of multilingual language processing in the brain. I suggest that one
avenue of research is to explore our current formulation of theoretical issues,
methodological strategies and forms of argumentation in neurolinguistics through
exemplars from the historical literature of our field.
References
Ahlsen, E. (2006), Introduction to Neurolinguistics. Amsterdam: John Benjamins.

Albert, M. and Obler, L. (1978), The Bilingual Brain. New York: Academic.
Atkin, K. and Lorch, M. (2006), ‘Reading without speech: Hyperlexia in a 4 year old
boy with autistic spectrum disorder’, Journal of Neurolinguistics, 19, 1–17.
Baron-Cohen, S. (2002), ‘The extreme male brain theory of autism’, Trends in Cognitive
Sciences, 6, 248–254.
Barrière, I. and Lorch, M. (2004), ‘Premature thoughts on writing disorders’, Neurocase,
10, 91–108.
Bastiaanse, R. and Grodzinsky, Y. (eds) (2000), Grammatical Disorders in Aphasia.
London: Whurr Press.
Bates, E. and Wulfeck, B. (1989), ‘Crosslinguistic studies of aphasia’, in B. MacWhinney
and E. Bates (eds), The Crosslinguistic Study of Sentence Processing. New York:
Bellugi, U., Lai, Z. and Wang, P. P. (1997), ‘Language, communication, and neural
systems in Williams syndrome’, Mental Retardation and Developmental Disabilities
Research Review, 3, 344–342.
Benson, D. F. (1967), ‘Fluency in aphasia: Correlation with radioactive scan localization’,
Cortex, 3, 373–394.
Benson, D. R. and Zaidel, E. (1985), The Dual Brain: Hemispheric Specialization in Humans.
New York: Guilford Press.
Breier, J., Maher, L., Novak, B. and Papnicolaou, A. (2006), ‘Functional imaging before
and after constraint-induced language therapy for aphasia using magnetoencepha-
lography’, Neurocase, 12, 322–331.
Caplan, D. (1987), Neurolinguistics and Linguistic Aphasiology: An Introduction. Cambridge:
Caplan, D. and Hildebrandt, N. (1988), Disorders of Syntactic Comprehension. Cambridge,
MA: MIT Press.
198 Marjorie Lorch
Cappa, S. F. and Vallar, G. (1992), ‘The role of the left and right hemispheres in recovery
from aphasia’, Aphasiology, 6, (4), 359–372.
Chiarello, C., Knight, R. and Mandel, M. (1982), ‘Aphasia in a prelingually deaf
woman’, Brain, 105, 29–51.
Chomsky, N. (1965), Aspects of the Theory of Syntax. Cambridge: M.I.T. Press.
Chomsky, N. (1981), Lectures on Government and Binding. Dordrecht: Foris.
Code, C. (1987), Language, Aphasia and the Right Hemisphere. Chicester: Wiley.
Comrie, B. (1981), Language Universals and Linguistic Typology: Syntax and Morphology.
Chicago: University of Chicago Press.
Devere, T., Trotter, J. and Cross, A. (2000), ‘Acute Aphasia in Multiple Sclerosis’,
Archives of Neurology, 57, 1207–1209.
Duffy, F. (1985), ‘The BEAM Method for neurophysiological diagnosis’, Annals of the
New York Academy of Sciences, 457, (1), 19–34.
Eling, P. (ed.) (1994), Reader in the History of Aphasia: From Franz Gall to Norman Geschwind.
Amsterdam: John Benjamins.
Fabbro, F. (2001), ‘The bilingual brain’, Brain and Language, 79, 201–222.
Gardner, H. (1999), Intelligence Reframed. New York: Basic Books.
Geschwind, N. and Galaburda, A. (eds) (1984), Cerebral Dominance: The Biological
Foundations. Cambridge, MA: Harvard University Press.
Geschwind, N. and Galaburda, A. (1985), ‘Cerebral lateralization: Biological mecha-
nisms, associations, and pathology: I. A hypothesis and a program for research’,
Archives of Neurology, 42, 428–459.
Goodglass, H. (1993), Understanding Aphasia. New York: Academic.
Goodglass, H., and Blumstein, S. (eds) (1973), Psycholinguistics and Aphasia. Baltimore:
John Hopkins University Press.
Goodglass, H. and Kaplan, E. (1976), Assessment of Aphasia and Related Disorders.
Philadelphia: Lea and Febiger.
Gopnik, M. and Crago, M. B. (1991), ‘Familial aggregation of a developmental
language disorder’, Cognition, 39, 1–50.
Green, D. (1986), ‘Control, activation, and resource: A framework and model for the
control of speech in bilinguals’, Brain and Language, 27, 210–223.
Green, D. (1993), ‘Towards a model of L2 comprehension and production’, in
R. Schreuder and W. B. (eds), The Bilingual Lexicon. Philadelphia: John Benjamins,
249–277.
Green, D. (1998), ‘Mental control of the bilingual lexico-semantic system’, Bilingualism
1, 67–81.
Grodzinsky, Y. (1990), ‘Theoretical perspectives on language deficits’, Cambridge, MA:
MIT Press.
Grosjean, F. (1982), Life with Two Languages. Cambridge, MA: Harvard University
Press.
Hellal, P. and Lorch, M. The Emergence of the Age Variable in Nineteenth Century Neu-
rology: Considerations of Recovery Patterns in Acquired Childhood Aphasia. In Fin-
ger, S. Boller, F. and Tyler, K.L. (eds.) Handbook of Clinical Neurology, Vol 95 (3rd series)
History of Neurology. Elsevier, Edinburgh, 2009, 845–52.
Hillis, A. E., Barker, P., Beauchamp, N., Winters, B., Mirski, M. and Wityk, R. (2001),
‘Restoring blood pressure reperfused Wernicke’s area and improved language’,
Neurology, 56, 670–672.
Hodges, J. R., Patterson, K., Oxbury, S. and Funnell, E. (1992), ‘Semantic dementia:
Progressive fluent aphasia with temporal lobe atrophy’, Brain, 115, (6), 1783–1806.
Hull, R. G. (2003), ‘How does bilingualism matter? A meta-analytic tale of two hemi-
spheres’. PhD Dissertation, Texas A&M University. from http://txspace.tamu.edu/
bitstream/1969.1/240/1/etd-tamu-2003A-2003032723-Hull-1.pdf, accessed 12.02.09
Jenkins, L. (2000), Biolinguistics: Exploring the Biology of Language. Cambridge: Cambridge
University Press.
Karmiloff-Smith, A. (1992), Beyond Modularity: A Developmental Perspective on Cognitive
Science. Cambridge, Mass: MIT Press/Bradford Books.
Kean, M. L. (ed.) (1985), Agrammatism. New York: Academic Press.
Kertesz, A. (1982), Western Aphasia Battery. San Antonio, TX: Harcourt Assessments.
Kimura, D. (1999), Sex and Cognition. Cambridge, MA: MIT Press.
Kushner, H. I. (1999), A Cursing Brain? The Histories of Tourette Syndrome. Cambridge,
MA: Harvard University Press.
Kushner, H. I. (2003), Crossing the Bridge: Toward an Applied History of Medicine.
The Academic Exchange, Atlanta: Emory University. from http://www.emory.edu/
ACAD_EXCHANGE/2003/octnov/hkushner.html, accessed 12.02.09
Lenneberg, E. H. (1967), Biological Foundations of Language. New York: John Wiley & Sons.
Lorch, M. (2006a), ‘Bilingualism and memory: Early 19th century ideas about the
significance of polyglot aphasia’, Cortex, 43, 658–666.
Lorch, M. (2006b), ‘Phrenology and Methodology, or “playing tennis with the net
down”’. Aphasiology, 20, 1059–1071.
Lorch, M. (2006c), ‘Language and memory disorder in the case of Jonathan Swift:
considerations on retrospective diagnosis’, Brain, 129, 3127–3137.
Lorch, M. (2008), ‘Multiple languages, memory and regression: An examination of
Ribot’s law’, Aphasiology, doi:10.1080/02687030801931182
Lorch, M. and Barrière, I. (2002), ‘Pitres’ two remarkable cases: pure agraphia (1884)
and polyglot aphasia (1895)’, in F. Fabbro (ed.), Advances in the Neurolinguistics of
Bilingualism: essays in honor of Michel Paradis. Udine: Forum Press, pp. 193–205.
Lorch, M. and Barrière, I. (2003), ‘The history of written language disorders: re-
examining Pitres’ case (1884) of pure agraphia’, Brain and Language, 85, 271–279.
Lorch, M., Ferner, R., Golding, J. and Whurr, R. (1999), ‘The nature of speech and
language impairment in adults with neurofibromatosis’, Journal of Neurolinguistics,
12, 157–165.
McNamara, P. and Durso, R. (2003), ‘Pragmatic communication skills in Parkinson’s
Disease’, Brain & Language, 84, 414–423.
Mechelli, A., Crinion, J. T., Noppeney, U., O’Doherty, J., Ashburner, J., Frackowiak,
R. S. and Price, C.J (2004), ‘Neurolinguistics: Structural plasticity in the bilingual
brain’, Nature, 431, 757.
Menn, L. and Obler, L. (eds) (1990), Agrammatic Aphasia: A Cross-Language Narrative
Sourcebook. Amsterdam: Benjamins.
200 Marjorie Lorch
Minkowski, M. (1927), Klinischer Beitrag zur Aphasie bei Polyglotten, speziell im

Hinblick aufs Schweizerdeutsche Schweizer Archiv für Neurologie und Psychiatrie, 21,
43–72.
Naeser, M. and Hayward, R. (1978), ‘Lesion localization in aphasia with cranial com-
puted tomography and the Boston Diagnostic Aphasia Exam’, Neurology, 28, 545.
Nicholas, M., Obler, L., Albert, M. and Helm-Estabrooks, N. (1985), ‘Empty speech in
Alzheimer’s disease and fluent aphasia’, Journal of Speech and Hearing Research, 28,
405–410.
Paradis, M. (1977). ‘Bilingualism and aphasia’, in H. Whitaker and H. Whitaker (eds),
Studies in Neurolinguistics (Vol. 3, pp. 65–121). New York: Academic Press.
Paradis, M. (ed.) (1983), Readings on Aphasia in Bilinguals and Polyglots. Montreal:
Didier.
Paradis, M. (ed.) (1998), Pragmatics in Neurogenic Communication Disorders. Oxford:
Elsevier.
Paradis, M. (2000), ‘The neurolinguistics of bilingualism in the next decades’, Brain
and Language, 71, 178–180.
Paradis, M. (2001), Bilingual and Polyglot Aphasia. Handbook of Neuropsychology (Second
Edition. Volume 3). Oxford: Elsevier Science, pp. 69–91.
Paradis, M. (2004), A Neurolinguistic Theory of Bilingualism. Amsterdam: Benjamins.
Paradis, M., Hagiwara, H. and Hildebrandt, N. (1985), Neurolinguistic Aspects of the
Japanese Writing System. New York: Academic Press.
Pick, A. (1921/1973), Aphasia [English trans. Jason Brown]. Springfield, IL: Charles C.
Thomas.
Pitres, A. (1895), ‘Etude sur l’aphasie chez les polyglottes. Revue de médecine 15,
873–899. Translated into English by M. Paradis’, in M. Paradis (ed.), Readings in
Aphasia in Bilinguals and Polyglots. Montreal: Didier, Vol. 15, 873–899.
Poizner, H., Klima, E. S. and Bellugi, U. (1987), What the Hands Reveal about the Brain.
Cambridge, MA: MIT Press.
Porch, B. (1971), Porch Index of Communicative Ability. Palo Alto, CA: Consulting
Psychologists Press.
Rapcsak, S., Krupp, L., Rubens, A. and Reim, J. (1990), ‘Mixed transcortical aphasia
without anatomic isolation of the speech area’, Stroke, 21, (6), 953–956.
Ribot, T. (1881), Les maladies de la mémoire. Paris: Alcan. Diseases of Memory, 1882.
(English translation of Les maladies de la mémoire, 1881.) London: Kegan Paul,
Trench & Co.
Schuell, H. (1965), Differential Diagnosis of Aphasia with the Minnesota Test. Minnesota:
University of Minnesota Press.
Schuell, H., Jenkins, J. and Jimenez-Pabon, E. (1964), Aphasia in Adults: Diagnosis,
Prognosis and Treatment. New York: Harper and Row.
Schwartz, M. F. (1984), ‘What the classical aphasia categories can’t do for us, and why’,
Brain and Language, 21, 3–8.
Shallice, T. (1988), From Neuropsychology to Mental Structure. Cambridge: Cambridge
University Press.
Tager-Flusberg, H. (ed.) (1994), Constraints on Language Acquisition: Studies of Atypical
Children. Hillsdale, NJ: Lawrence Erlbaum.
Vargha-Khadem, F., Gadian, D. G., Copp, A. and Mishkin, M. (2005), ‘FOXP2 and the
neuroanatomy of speech and language’, Nature Reviews Neuroscience, 6, 131–137.
Weisenburg, T. and McBride, K. E. (1935), Aphasia. New York: Commonwealth Fund.
Whitaker, H. and Whitaker, H. A. V. (eds) (1976–1979), Studies in Neurolinguistics
(Vol. 1–4). New York: Academic Press.
Whurr, R., Lorch, M., Fontana, H., Brookes, G., Lees, A., and Marsden, C. D. (1993),
‘The use of botulinum toxin in the treatment of adductor spasmodic dysphonia’,
Journal of Neurology, Neurosurgery and Psychiatry, 56, 92–111.
CHAPTER
11
Clinical Phonology
Martin J. Ball and Nicole Müller
11.1 Introduction
Clinical Linguistics might be thought of as a classic example of the use of linguis-

tics in the real world, where the interaction of linguistic theory and clinical data
allows a classification of disorder type, an analysis of disorder severity and a path-
way into principled remediation. In this chapter, however, we argue that Clinical
Phonology, as currently described, may be suffering from a misunderstanding
of the role of much of theoretical phonology. The current dominant approaches
to phonological theory are overtly not attempts to model the act of speech produc-
tion but rather attempts to provide elegant descriptions of data. This appears to be
sometimes misunderstood by those applying these models to disordered speech,
however, in that the descriptive frameworks afforded by these frameworks come
to be treated as models of speech production. Further, recent research into
covert contrasts is beginning to suggest that, with child speech disorders at any
rate, errors perceived as phonological may in many cases have a motor cause. This
leads to speculation as to how useful the traditional distinction between phonetic
and phonological disorders is, and also how useful this distinction is at the level of
intervention.
We argue here that we need to move towards a phonological description that
overtly does attempt to model speech production, and that this provides a more
useful analysis of clinical data, and a better principled direction for remediation.
11.2 Clinical Phonology
The application of insights from theoretical phonology to the analysis of disordered

speech has been a feature of clinical linguistics for a long time (see Grunwell, 1987,
for an overview of early work), and might be thought of as one of the more fruitful
collaborations between linguistics and communication disorders. Of course, theo-
retical phonology has not remained static since the major breakthrough of
Chomsky and Halle’s 1968 publication, The Sound Pattern of English (henceforth
SPE), which introduced generative phonology.
Indeed, theoretical phonology has been one of the most productive areas of lin-
guistics over the past half century in terms of both the development of the standard
Clinical Phonology 203
SPE model, and in a proliferation of alternative approaches. Advances from the clas-
sic SPE model include developments with features, e.g. feature geometry (Clements,
1985), and underspecification (Archangeli, 1988) and developments with derivation
(cyclical rules and lexical phonology: Kiparsky, 1982a, b, 1985; Kenstowicz, 1994).
A major shift in thinking brought about the idea of non-linear phonology, with
work in autosegmental and metrical approaches (see the overview in Goldsmith,
1990). Later, some phonologists moved away from traditional binary (or mixed
binary and unary) feature systems, developing theories such as dependency phono-
logy (Anderson and Durand, 1986, 1987; Anderson and Ewen, 1987), government
phonology (Kaye et al., 1985, 1990; Harris, 1990, 1994) and radical CV phonology
(van der Hulst, 1995, 1996). Despite all this work, however, it is often unclear
whether these developments in theory are attempts to produce a more elegant way
to describe sound patterns, or whether they are intended to model the neurolin-
guistic and/or psycholinguistic aspects of speech production. This dichotomy we
refer to as the difference between a descriptively adequate and psycholinguistically ade-
quate phonology.
These terms are developments of those used in descriptions of linguistic models
and were coined by Noam Chomsky. For example, Chomsky (1965) proposed that
a linguistic theory has observational adequacy if it accounts for linguistic behav-
iours in as elegant and non-redundant a way as possible, and descriptive adequacy
if it accounts for the linguistic intuitions of speakers in addition to generating
all and only the structures of a language. A theory has explanatory adequacy if it
can account for why one particular descriptively adequate grammar can be pre-
ferred over another. For our purposes here, we suggest a binary division between
descriptive and psycholinguistic adequacy is all that is needed. For psycholinguistic
adequacy a theory should attempt to model how the brain organizes particular
linguistic behaviours and operationalizes them. Such a division in types of ade-
quacy applies to models of phonology just as much as to other areas of linguistic
organization.
Early Generative models of phonology (e.g. SPE) strove to achieve descriptive
adequacy, but did not claim the sort of psycholinguistic adequacy we described
above. Indeed, Chomsky (1972: 117) himself notes this point:
[A]lthough we may describe the grammar G as a system of processes and rules

that apply in certain order to relate sound and meaning, we are not entitled to
take this as a description of the successive acts of a performance model . . .
Developments such as autosegmental and metrical phonology (Goldsmith,

1990), and feature geometry (Clements, 1985), can all be seen as attempts to pro-
duce more elegant descriptions of phonological phenomena. Nevertheless, even
within the Generative school, approaches such as underspecification (Archangeli,
1988) might be justified by appeals to psycholinguistic plausibility. It was argued,
e.g., that underspecified segments would take less memory storage space than
fully specified ones (see Spencer’s 1996 comment in this regard, p. 305, n. 7). In
application to disordered speech, too, the concept appears to have acquired a
204 Martin J. Ball and Nicole Müller
psycholinguistic aspect: ‘children’s substitution patterns can be seen as coming

about from their underspecifying underlyingly two or more sounds that are con-
trastive in the target system with that feature being filled in by default’ (Dinnsen
and Barlow, 1998: 71).
Other phonological approaches have more clearly identified themselves as
models of speech production (some psycholinguistic, others motor-based). These
include natural phonology, and articulatory or gestural phonology. Stampe
(Donegan and Stampe, 1979) argued in favour of a ‘Natural Phonology’ that was
explicitly innatist: in other words a search for psycholinguistic adequacy. Articu-
latory (or Gestural) Phonology (Browman and Goldstein, 1986, 1992) is based on
articulatory gestures (and so ultimately motor programs), Government Phonology
(Harris and Lindsey, 1995, 2000) on acoustic patterns and Sonority theory (Yavaş,
2003) on perception. The current focus of interest in theoretical phonology,
Optimality Theory, is treated in clinical phonology as though based on psycho-
linguistic principles. We shall discuss this approach further below.
11.3 Desiderata for a Clinical Phonology
When we consider what we need a clinical phonology to do, we encounter again

the dichotomy between the description of data and the modelling of behaviours.
Let us first consider the description of data:
1. A clinical phonological analysis should acknowledge that the starting point in

the clinical endeavour is the analysis of speech output, and that the data used
are comprehensive, naturalistic and therefore representative of the client.
2. A clinical phonological analysis should be able to group together patterns of
usage (i.e. patterns connected to particular sound classes) in a principled way,
using a formalism that is elegant and economical.
3. A clinical phonological analysis should aim to describe patterns in a client’s
speech output first of all in terms of the data themselves. Systematic relation-
ships (such as contrasts) should be allowed to emerge from the data. In addi-
tion, it is necessary to describe relationships between a client’s realizations and
target patterns, and to show whether any patterns uncovered are subject to vari-
ation (either contextual or stylistic).
4. A clinical phonological analysis should incorporate some kind of metric that
allows an evaluation between the client’s phonology and that of normal speak-
ers, other clients and the same client at different stages in therapy.
Now we can turn attention to the modelling of speech behaviours:
1. A clinical phonology should be able to distinguish between speech output prob-

lems that result from a disruption to the client’s phonology, and those that result
from disruptions to the motor planning or motor execution of an intact phono-
logy (for the time being, we adopt this tripartite division, without any claims as
to the nature of a ‘phonology’, or the nature of its representation or prime).
2. A clinical phonology should be able to distinguish between the source of a dis-

ordered realization, and the effect that has on the listener (see further below).
3. A clinical phonology should be able to model the speech production (and pos-
sibly also perception) process using a psycholinguistically adequate account.
4. A clinical phonology should inform therapeutic intervention based on a psy-
cholinguistically adequate metric of the disorder and psycholinguistically rele-
vant plan.
As can be seen, these desiderata do not always mesh together cleanly: An account
of the data that is both elegant and economical might not also be psycholinguisti-
cally plausible (see below for a more detailed discussion), and a metric derived
from a descriptively adequate phonological analysis might not be the same as a
psycholinguistically adequate metric, and, finally, descriptively adequate accounts
may be unable to distinguish between source and effect differences in clinical data.
11.4 Phonology (and Phonetics) in the Clinic
It is a curious circumstance that the long history of reflection and debate about the
nature of phonology, its representation and its prime (or smallest unit in phono-
logical structure) has had very little influence on clinical practice, with two excep-
tions: Phoneme Theory, and Natural Phonology and process analysis (we shall
return below to a discussion on the application of other theoretical approaches to
clinical data; these have, however, not achieved wide currency in clinical practice).
Our discussion in this section and elsewhere in this chapter is situated in a specific
context of culture, namely that of speech-language pathology education and prac-
tice in the United States, since this is the context in which we have chiefly been
operating over the past years.
Clinical practice, and literature concerned with clinical practice show evidence
that there is no systematic or consistent separation of speech output as opposed to
phonology. The preferred prime of analysis is the ‘phoneme’, and even though
course texts aimed at articulation and phonological impairments tend to point out
that phonemes are theoretical entities, and speech sounds physical entities (e.g.
Bauman-Waengler, 2004), ‘phonemes’ are often implicitly treated as ‘real’.
A recent survey on assessment practices used by speech-language pathologists
working with children with suspected speech sound disorders (Skahan et al., 2007)
revealed that the most commonly used published tests are the Goldman-Fristoe
Test of Articulation (GFTA; second edition: GFTA-2, Goldman and Fristoe, 2000),
and the Khan-Lewis Phonological Analysis (KLPA; second edition: KLPA-2, Khan
and Lewis, 2002). The most commonly employed analysis procedure used by
clinicians was a phonological process analysis. GFTA-2 is a test of articulation, i.e.
of speech output, using single word picture stimuli and two brief stories to elicit
consonant sounds in different places in word position, as singletons and in clusters.
Analysis is in terms of the phoneme contrasts of Standard American English, as
is the evaluation of results in comparison to developmental matrices. There is
thus no conceptual separation between the nature of phonological contrast, and
‘phonemes’, and speech output. KLPA-2 uses the output of GFTA-2 in order to
identify error patterns in a client’s speech output, formulated in terms of phono-
logical processes. Of the patterns that are provided for in KLPA-2, ten are scored
on the assessment, whereas the others are not. KLPA-2 makes no reference to the
theory of Natural Phonology from which the term ‘process’ was borrowed (Stampe,
1979; Donegan and Stampe, 1979), and which in essence is an attempt to map
the commonalities found in the normal acquisition of language, both across indi-
viduals and across languages, in terms of mental operations affecting speech out-
put. Thus the term process has become divorced from its theoretical origins,
and has come to be used synonymously with ‘pattern’. We shall return below to the
‘de-theorization’ of constructs from natural phonology, especially the notion of
the process.
GFTA-2 and KLPA-2 exemplify a widespread, essentially atheoretical approach
to speech sound disorders, their assessment and remediation. As Kamhi (2006:
275) sums up with reference to intervention methods, ‘[m]any clinicians are not
wedded to a particular theoretical orientation, or they indicate that they are eclec-
tic and use whatever works.’ Thus speech sound errors are not ‘theorized’ in clini-
cal practice in terms of, e.g., speech output on the one hand and the nature of
underlying systems or representations on the other, or the effects of speech output
on the listener, and the nature of the producing systems. The terminological usage
of researchers writing for and about clinical practice mirrors the recasting of
theoretical into real entities, as in e.g. Skahan et al. (2007: 253), who state that
‘[d]ocumenting a phonetic inventory is useful for describing the child’s unique
sound system and the phonemes the child is capable of making.’ Similarly, Kamhi
(2006: 273) states that ‘[a]lthough individual sounds may not exist in neatly pack-
aged units, there has been considerable interest in the age at which children are
able to produce each of the phonemes correctly in their language.’
The ‘phonological revolution’ of the decades leading up to the 1980s has led to
a situation in clinical practice where, essentially, phonetics and phonology are not
operationalized as distinct entities (see also Shriberg and Kwiatkowski, 1982a, b,
for further discussion of labelling of speech sound disorders). The driving princi-
ples of analysis in clinical practice appear to be, first of all, the reality of speech
output. Second, the existence and reality of the phoneme-sized contrastive unit is
tacitly assumed, and is the kingpin of analysis in terms of target systems, which
again are operationalized in terms of inventories of phonemes (this leads to an
analysis of distortions versus substitutions, which is discussed further below). A
further driving principle is the search for patterns in the speech output, which are
tacitly equated with patterns in underlying systems, and most frequently expressed
by process labels. One could look at this as nothing more than a pragmatic
approach to the complexity of speech output problems. After all, the starting point
of any analysis is a client’s speech output, and our measure of success is again the
improvement of said output. However, the metaphors and categories we use of
course constrain our view of our data. Our discussion of some common error pat-
terns will show that a failure to theorize speech output problems, their origins and
their effects, can obscure important patterns.
The commonly used differentiation between distortion errors and substitution

patterns arises out of the analysis of a client’s speech output exclusively in terms
of the target system. In other words, if a client produces, instead of a target sound,
a sound that is phonetically related to the target, but does not map onto another
contrastive sound in the target system, an error is treated as a ‘distortion’, and by
implication assumed to be an issue of production, with an intact underlying target.
The realization of target /r/ as [] would be an example. If, on the other hand, a
client’s realization maps onto another unit in the target system (such as a client’s
production of [w] for /r/), an error is treated as a substitution, and by implication
as an issue of the client’s sound system.
Indeed, the distinction between ‘distortions’ and ‘substitutions’ has been criti-
cized by several authors. Grunwell (1985a, b, 1987, 1988) has commented on the
inadequacy of these terms for the classification of speech errors. She has made it
clear that often these two categories are insufficient for the analysis of her data,
especially when sounds that are not within the target phonology are used. Harris
and Cottam (1985), Hawkins (1985), Hewlett (1985) all contributed to a debate
on the phonetics-phonology distinction where it became clear that there was a
strong case for disentangling the ‘cause’ and the ‘linguistic effect’ of disordered
speech. In other words, as illustrated in Ball and Müller (2002) we may have a pho-
nological cause with an articulatory effect, or an articulatory cause with a phono-
logical effect. The /r/ problem noted above is an illustration of this: we may well
posit that the frequency of /r/ problems in English is due to an articulatory diffi-
culty (e.g. fine motor control of the tongue apex in the case of apical []); one
common realization ([w]) results in a phonological disruption to the client’s
speech in terms of its effect on the listener, the other ([]) does not.
The contribution of Grunwell and others to the analysis of error patterns,
regarded as phonological rather than phonetic, has to be acknowledged. In clini-
cal application, as outlined above, a re-focusing of speech output problems as
phonological (see e.g. Shriberg and Kwiatkowski, 1982a, b) has eventually led to a
blurring of all distinctions between speech output, and the mechanisms (psycho-
logical and physiological) contributing to it. However, a clinical phonology has
to integrate issues of representation of speech, motor planning and execution
(or articulation). A brief examination of some commonly occurring error patterns,
which, in the Grunwellian tradition are typically summarized as phonological proc-
esses, will illustrate this point.
1. Velar fronting. This process sees target velar consonants realized as alveolars
(or other anterior consonants depending on the target language), and could
be thought of as an archetypal phonological error in that a whole phonological
place category is missing and the replacements are other phonemes of the
language and thus contrastive function is impaired. However, work on ‘covert
contrasts’ by Gibbon and her colleagues (see Gibbon, 1990, 1999, 2007; Gibbon
and Scobbie, 1997; Gibbon and Wood, 2001) shows that this superficially
simple process is more complex. Gibbon has found that covert contrasts may
well exist that are not obvious impressionistically. Instrumental techniques
(e.g. electropalatography) have shown that apparent mergers of alveolars with

velars are often produced with significantly different tongue-palate contact
patterns. So, what appears to the listener to be a phonological merger, is main-
tained by the speaker as phonetically different.
2. Fricative stopping. Again, the effect of this pattern is to remove a whole cate-
gory of target consonants from the child’s inventory, with major consequences
for the ability to signal phonological contrasts. However, the question we need
to ask is why is fricative stopping so common both in child disordered speech
and normal development, but plosive spirantization so rare? The underlying
cause of fricative stopping is unlikely to be purely phonological (e.g. that there
are too many units in the system, or that some abstract constraint needs to
be relaxed), but a motor cause: the requirement for fine motor control of
the articulators to produce the necessary narrow air channel for frication as
opposed to the grosser motor control needed to make a stop closure.
3. Liquid gliding. The commonest found form of this (problems with /r/) was
referred to above. The less common realization of /l/ as [j] can be considered
equally to have an articulatory origin: the difficulty in the fine motor control of
the tongue rims.
4. Context sensitive voicing. At first sight, context sensitive voicing does seem an
ideal candidate for a phonological explanation: voiced and voiceless obstruents
are both produced, but not at the same place in structure (voiced only
syllable-initially, voiceless only syllable-finally). However, word final devoicing is
commonly encountered in natural language. Many Germanic languages, for
example, only allow voiceless obstruents word-finally, and even in English there
is considerable devoicing in word-final position, with the contrast maintained
through strategies such as vowel clipping (shortening of vowel duration before
voiceless obstruents). This can be explained aerodynamically, as it is difficult to
maintain sufficient subglottal pressure for phonation as one nears the end of
an utterance and airflow is expended. Conversely, it is easier to use phonation
utterance initially when subglottal pressure is strong.
5. Cluster reduction. Again, it would appear that this pattern is a good candidate
for a phonological explanation. Both consonants of a binary cluster may well be
found in the client’s speech as singletons, but not as clusters. So, does this pat-
tern derive from a phonology lacking the ability to have two (or more) conso-
nants in a single place in word structure? Evidence from both normal and
disordered speech (e.g. Gierut, 1999) suggests that here too there is an alterna-
tive, motor explanation. Apart from the total absence of one member of a tar-
get cluster, we may also see feature synthesis (or coalescence) and the insertion
of epenthetic vowels. So, e.g., a target /sl-/ may be realized as [ŵ-] (with the
voiceless friction of the /s/ being merged with the alveolar laterality of the /l/).
A target /tr-/ may be realized as [t-] with a schwa intervening between the
two consonants of the cluster. Such realizations show a gradual movement
towards a cluster using strategies that produce a simpler motoric task, and
retain the canonical CV structure that Öhman’s (1966) ground-breaking work
posited as the basic articulation type. So it is arguable that cluster reduction
is not the absence of an organization category of cluster, but the inability to
produce two consonants on the consonant parameter (as described by Öhman,

1966) between the vocalic elements produced on the vowel parameter.
What we suggest here, therefore, is that many child speech error patterns
typically described as phonological, are only phonological if clinical phonology
describes the effect of the error, in the sense that a contrast or group of contrasts
is absent from speech output as perceived by a listener. However, these error pat-
terns often seem to have a motor (or articulatory) origin.
We do not mean to suggest, however, that there are no phonological errors to
be encountered in the clinic (using phonological here to refer to the cause rather
than the effect). Many adult acquired neurological disorders exhibit speech disor-
ders that cannot be accounted for simply in articulatory terms. Ball and Müller
(2002) argue (following others such as Hewlett, 1985; Code and Ball, 1988) that
we need a tripartite division of speech errors: paraphasia type errors would impli-
cate the phonological representation as the primary site of the disorder, apraxia
errors would be mainly at the phonetic planning level, while dysarthric errors
would be situated at the articulatory level. Kent (1996) has demonstrated that we
need to be aware of the interactions between different levels when considering
these different types of impairments. We return below to consider these levels and
a model of phonology in use that avoids a level of abstract phonology and com-
bines paraphasia type errors and apraxia type errors because phonological storage
contains both the physical gesture and the coordination of gestures.
Dodd (1995) and Bradford and Dodd (1995) have also questioned whether
child speech disorders are ultimately motor based. They concluded that two of
their four groups of subjects did display motor deficits in undertaking various
motor tasks, but that two did not (and are thus described as having phonological
disorders). However, we should note that the tasks undertaken by the subjects were
mostly non-linguistic, and non-naturalistic. Also, the authors used a binary distinc-
tion between an abstract phonological level and a motor level (where we have
argued for a three-way division1). Further, the analysis was in terms of groups of
children and did not examine the patterns of speech problems themselves which,
we have just argued, often show an articulatory cause or covert contrasts.
Most recently, Gibbon (2007) has argued, from the evidence of her own and
others’ work that the reliance on phonological explanations for much disordered
speech needs to be challenged. She notes, ‘data from instrumental studies reveal-
ing phenomena such as covert contrasts and undifferentiated gestures cast some
doubt on these conclusions, suggesting that subtle phonetic difficulties could
underlie many of the surface patterns that we hear in the speech of children with
phonological disorders’ (Gibbon, 2007: 254).
11.5 Current Phonological Theory and Clinical Data
As we noted earlier, clinical phonology researchers have applied many models of

theoretical phonology to clinical data. The current dominant model of phonology –
Optimality Theory (or constraint-based phonology, see Prince and Smolensky,
1993; Archangeli and Langendoen, 1997) – is another example of a theory that is
in search of the economy and elegance that is the metric of descriptive adequacy,
but one where some authors also imply a psycholinguistic power to its descriptions.
For example, in the application of this model to disordered speech Dinnsen and
Gierut (2008) suggest that ‘The clinical significance of this point [constraint rank-
ing] is that treatment can be designed to focus on the demotion of that top-ranked
constraint, bringing about the suppression of that and other related error patterns
without directly treating those other error patterns.’ The demotion of the top-
ranked constraint presumably happens in a client’s mental phonology. Yet, at the
same time, the theory relies on an infinite number of possible pronunciations to
be scanned by the evaluation (Eval) module, which as a model of speech produc-
tion would surely result in no speech ever being produced. McCarthy (2002: 10)
addresses this last point directly:
how can Eval sort an infinite set of candidates in finite time . . . ? The error lies
in asking how long Eval takes to execute. It is entirely appropriate to ask whether
Eval . . . is well defined, captures linguistically significant generalizations, and
so on. But questions about execution time or other aspects of (neural) computa-
tion are properly part of the performance model PM and must be addressed
as such.
Under this reading, therefore, we must conclude that Optimality Theory (OT) is
not intended as a psycholinguistic model of speech production.
We have argued above that much of what is described in child speech disorders
as impaired phonology, may well be better characterized as originating in articula-
tory or motor difficulties. So, where does this leave clinical phonology? Whether
we use processes, constraints or rules, if we stick to abstract phonological labels we
will not be capturing the fine phonetic detail of much disordered speech. For
example, the client who realizes word-initial English /r/ as a velar approximant
(Müller et al., 2008) will fit into no natural process, and will require an ad hoc
constraint to characterize their speech. We turn in our final section to possible
ways forward.
11.6 What We Might Move To
As we have discussed above, there is essentially a disconnect between current

work in theoretical phonology applied to clinical data, and phonology as used in
clinical practice. The descriptive categories of the dominant paradigm in theoreti-
cal clinical phonology (Optimality Theory) have, in the course of clinical applica-
tion, come to be regarded at least by some authors as representing psychologically
realistic entities. In much of clinical practice, the question as to the nature of
the interaction between mental and motor events, in other words, the nature of a
phonology as it relates to phonetics, is not asked. Neither approach appears to us
the most promising.
One way forward might be to move to a surface description that is more phonetic
than most phonological approaches, such that errors that appear phonological and
those that appear articulatory can be accounted for by the same framework. An
obvious candidate is Gestural Phonology (Browman and Goldstein, 1986, 1992).
Gestural, or Articulatory, phonology addresses the problem of phonological organ-
ization from a phonetic perspective, and proposes the notion that ‘. . . phonology
is a set of relations among physically real events’ (Browman and Goldstein, 1992:
156). These real events are called ‘gestures’ in articulatory phonology and consti-
tute the prime of the theory. The relations between the gestures, and their relative
timing can be diagrammed on gestural scores, and the covert contrast between a
target alveolar and a target velar sounding like an alveolar (see discussion in velar
fronting earlier) can be clearly shown, as in Figures 11.1 and 11.2.
h u
t
VELUM
TT closed alveolar
TB
narrow velar
LIPS open lab
GLOTTIS wide
Figure 11.1 Gestural score for target alveolar in client showing ‘velar fronting’.
h
t u (> khu)
VELUM
TT closed alveolar
TB
closed velar narrow velar
LIPS open lab
GLOTTIS wide
Figure 11.2 Gestural score for target velar sounding like alveolar in client showing ‘velar
fronting’ (showing covert contrast).
Such an approach provides an elegant surface description as well as showing

underlying causes that may differ in terms of surface effects (thereby meeting
several of the desiderata we described earlier). A deeper psycholinguistically ade-
quate account of disordered speech might be available were we to embed the
gestural formalism within a cognitive approach to phonology. Various approaches
to psycholinguistic modelling of speech production and perception are reviewed
in Baker et al. (2001), but the model we will examine here is more firmly based
within the traditions of phonology than many of those discussed there.
Ball (2003) and Sosa and Bybee (2008) have both described how Bybee’s (2001)
model of phonology in use could be applied to clinical data. This approach is
explicitly designed to model phonology as a cognitive activity. A cognitive phono-
logy2 account deals with issues of storage of items, establishment of networks
linking phonologically similar items and the emergence of units (such as contras-
tive units, or ‘phonemes’) from these networks. It posits highly redundant multi-
ple storage, and thus does not seek the economy of description that traditional
approaches do. We will briefly describe the main parts of this approach to what
Bybee (2001) called ‘phonology in use’.
11.7 A Cognitive Phonology
Frequency of use is stressed as an important aspect of Bybee’s (2001) model. She

notes that experience affects representation; therefore high frequency forms and
phrases have stronger representations in memory, and thus are resistant to ana-
logic change. Low frequency forms are more difficult to access, and therefore may
be subject to change or loss. Patterns that apply to more items are also stronger
and more productive.3
Cognitive Phonology also claims that that mental representations of linguistic
objects do not have predictable properties abstracted away, but are based on the
categorization of actual tokens. Such a claim is completely opposed to the accepted
wisdom of generative models, where predictable properties of units are abstracted
away via the process of derivation. Cognitive Phonology does not object to redun-
dant storage; indeed, it claims that this is in fact what happens. Generalizations
over forms are not separate from the stored representations of forms, according
to the cognitive approach, but emerge directly from them. Generalizations are
expressed as relations among forms based on phonetic/semantic similarities. So,
multimorphemic words are stored whole in the lexicon (i.e., non-derivationally).
Further, Cognitive Phonology states that categorization is based on identity or
similarity. Categorization organizes the storage of phonological percepts, as we will
exemplify below. Units such as morpheme, syllable, phoneme/segment are not
basic units of the theory, but are emergent: they arise from the relations of identity
and similarity that organize stored units. Therefore, the traditional distinction
between phonetic and phonological information is not part of storage in this
model; phonological contrastivity (in traditional terms) is emergent. As well as
storing units such as words, motor-articulatory plans to execute speech and acoustic-
perceptual ones to interpret it are also stored. Storage in Cognitive Phonology is
highly redundant (as opposed to the attempts at descriptive elegance of other
approaches), so schemas may describe the same pattern at different degrees of gen-
erality. This model, then, is not looking for descriptive economy (it is deliberately
highly redundant), but for psycholinguistic plausibility.
11.7.1 Organized Storage
Generative models claim that material appearing in rules cannot appear in the
lexicon and vice versa; this underlies the whole principle of derivation. Cognitive
Phonology explicitly rejects this, and claims that predictable features do appear in
schemas (patterns, rather than rules), but also in the lexicon. This means that
words and phrases can be stored whole, yet still participate in the schemas that
link similar forms. This is because storage is not linear but spatially networked.
Networks can link morphophonemic forms (e.g. the -ing ending on verbs in
English), or purely phonological (such as the /-/ in send, bend, mend). The
strength of these networks is important and derives in part from frequency of use.
Units such as phonemes and morphemes emerge from these schemas. The strength
of connections also depends upon the degree of similarity, and this point is
explored more fully in Bybee (1985).
So, what in generative models are derivational rules, in Cognitive Phonology are
schemas formed by links between forms in the lexicon. Cognitive Phonology is,
therefore, very ‘flat’ in terms of derivation, but multi-dimensional in terms of asso-
ciations, as virtually all realizations are stored lexically. Nevertheless, Cognitive
Phonology, as described by Bybee (2001) does recognize processes of simplifica-
tion. Two main types of processes are recognized: fast speech processes (lenitions),
and historical change (analogy or lenition). The effect of analogy can be blocked
by high-frequency use, whereas high-frequency use encourages lenition. Note,
however, that frequency of use does not play a part in the processes of fast speech
forms.
The tripartite analysis of speech disorders into phonological, motor planning
and motor implementation discussed earlier can be fitted into Bybee’s (2001)
model of phonology in use. Schemas, whether they have contrastive function or
not, are all within the phonology as organized storage. Thus gestures and their
coordination (i.e. motor planning) are all part of storage. Motor execution, how-
ever, would be independent of phonology. We await further research to inform us
in what ways the tokens and their motor plans are stored and linked. As meaning
is emergent in this approach, there is nothing to prevent an analyst noting the
meaning consequences of a particular error, however.
If we apply this approach to phonology to disordered speech, we would look for
both the descriptive aspects we have discussed and the explanatory ones (i.e. both
effect and cause).
11.7.2 Descriptive
If units such as phonemes and syllables are emergent, and processes are either
lenition or analogy, then we would need to describe the data differently in a Cogni-
tive Phonology approach than in other popular clinical assessment types. We would
need to describe data in terms of weak or non-existent lexical networks first, lead-
ing to specific problems with segments or other units. When patterns of lenition
are observed (such as in dysarthric speech) the formalisms of gestural phonology
can be employed in descriptions of the clinical data (Bybee, 2001, supports a
gestural approach to description). However, some child phonological disorders
appear to be more readily described in terms of fortitions (such as fricative
stopping). Bybee’s model does allow for different processes of simplification in
child phonological development, and it would appear that these would be needed
to account for commonly occurring patterns. These patterns, of course, can be
also be accounted for through the same gestural formalisms.
11.7.3 Explanatory
If there is inadequate lexical storage, leading to networks that are not established
strongly enough and not linked densely enough, then contrastive units (phonemes/
morphemes, etc.) will not emerge. So, we can assume that problems arise if insuffi-
cient items are stored correctly in the lexicon, or if the categorization that creates
connections does not occur. Of course, it may be difficult to know whether lexical
storage is faulty because there simply are not enough items stored to create the
proper networks and allow units to emerge, or because some or many items that
are stored, are stored incorrectly, thus disrupting expected networks, or whether
there is some problem linking tokens to their motor plans. This last point would
fit well with our argument that much child disordered speech results from articula-
tory difficulties.
What caused the incorrect storage or linking problems is, of course, a further
step in the interpretive process. Bybee (2001) notes that a plausible explanation of
historical changes that affect low frequency forms first could well be the result of
incorrect perception of the form during acquisition. As these forms are rarely
heard, there are few examples for the child to correct the misperception. The
other main source of sound change is rearrangement of articulatory gestures (usu-
ally lenition). This affects high frequency forms most, as there is the greatest bene-
fit to the user to simplify forms that are used most often. In disordered speech,
therefore, we may see incorrect storage due to perceptual breakdown or due to
articulatory difficulty (or a combination of the two).
11.7.4 Planning Remediation
In Bybee’s (2001) account of Cognitive Phonology, frequency of use is much more

important than notions of contrast. Therefore, in remediation, minimal pairs
drills, the staple of many other approaches to remediation that do stress contrast,
would not, it is assumed, be deemed of major importance. More important would
be using sets of words that would reinforce networks and allow specific units to
emerge. It could possibly be that a type of maximal opposition therapy (e.g. Gierut,
1990) that moves away from minimal pairs is successful because it does in fact
reinforce networks rather than concentrating on contrast. Clearly, this is an area
where further research is warranted.
11.8 Examples
Two brief examples, one child and one adult, can give an idea of how the model
we suggest here might work.
Ball et al. (2003) and Ball (2003) describe a 6-year-old girl with highly unintelli-
gible speech, one aspect of which was the use of double syllable onsets during syl-
lable repetition tasks. These double onsets consisted of a prolonged first consonant,
a slight pause and then a second consonant leading into the remainder of the syl-
lable. In some instances the first and second onset consonants were the same, in
other cases they were not; the first consonant, however, was usually the expected
form of the nonsense syllable she was attempting to copy. While the client did not
use these double onsets in connected speech, her patterns of target consonant
realizations did often mirror the choice of the second consonant.
Examples of these double onsets were:
z Repeat onset: [ - ] (target //)

z Disconnect nasality: [ - ] (target //)
z Disconnect manner and place: [ - ] (target //)
The repeat onset strategy was used for target /h/, and sometimes for target
/m/, /n/, /l/ and /r/, and occasionally for target /s/ and //. Disconnect nasal-
ity was the commonest strategy with /m/ and /n/, while disconnect manner and
place was commonly used with both the fricatives and sometimes with the liquids.
The only plosive tested (/t/) did not show the double onset strategy and was real-
ized correctly. Other English consonants were not included in this testing.
In terms of gestural phonology, we can see all three of these patterns as exam-
ples of lenition, as they involve the removal of gestures and thereby the simplifica-
tion of the gestural score. Figure 11.3 illustrates the gestural score for the syllable
structure in [ – ], where this simplification is clearly shown.
This client’s speech output was possibly a result (at least partly) of her medical
history in that disruptions to hearing and attention that were recorded could have
resulted in incorrect storage due to perceptual disruptions. Her malocclusion and
temperomandibular problems may also have resulted (at least at an early stage)
in difficulties with articulation which themselves may result in incorrect storage.
The notion of density of storage suggested by Bybee’s work is also seen in the
fact that the client could produce target consonants when prompted (i.e. she was
s -  u
VELUM
TT crit alv
TB narrow velar
LIPS open lab
GLOTTIS wide closed
Figure 11.3 Gestural score for [s–u]. (From Ball et al., 2003.)
‘stimulable’ for these sounds), but rarely in spontaneous speech. This implies that
these forms are stored, but that substitution forms have stronger links and so are
accessed more readily. The dominant consonant in her connected speech was a
glottal stop, i.e. the ultimate lenited form. These lenited forms are also motorically
simpler, thus fitting with our earlier argument.
Although these substitution patterns in a traditional analysis might suggest that
the client had a mix of phonological and phonetic disruptions, a gestural analysis
coupled to a cognitive model of storage reveal the essential unity of the patterns
displayed in the double onset examples.
Ball et al. (2004) and Code et al. (2006) describe a case of progressive speech
degeneration in a 63-year-old male. Among the characteristics of his speech
described are: reduction of Voice Onset Time in fortis plosives over time leading
to loss of aspiration, variable change of nasal stops to oral stops, frequent use of
glottal stops, insertion of epenthetic vowels in final clusters; addition of [] to words
ending in /u/ and addition of a labial obstruent to words beginning with /r/.
Following Bybee’s (2001) assertion that lenitions account for the majority of pho-
nological processes, we assign the first three of these patterns to that category.
As noted immediately above, we do this under a gestural understanding of leni-
tion, whereby (rather than using the traditional sonority metric) we judge lenition
to involve either the reduction in the number of gestures involved in a gestural
score, or the opening of the constriction degree of a gesture. Figure 11.4 shows the
realization of ‘more’ as [], and the gestural score clearly illustrates the reduction
in the number of gestures.
The remaining errors listed above involve the decoupling of gestures (epen-
thetic vowels in clusters, and labial obstruents added to /r/ initial words), and for
the addition of dark-l to /u/ final words we see the addition of a tongue-tip gesture
with the maintenance of the tongue body location through the back vowel into the
m c → b c
VELUM wide
TT
TB narrow uvular narrow uvular
LIPS closed labial open labial closed labial open labial
GLOTTIS
Figure 11.4 Gestural scores for the nasal and denasal realizations of ‘more’. (From Ball
et al., 2008.)
dark-l. It is difficult to claim that these decoupling and reorganizing of gestures are
also lenitions. As Bybee allows for different processes (e.g. fortitions) in acquisi-
tion, we also need such flexibility in disordered speech. However, we have to ask
whether these errors in this case are actually phonological.
Earlier in this chapter we discussed a tripartite approach to sound production:
phonological organization, motor planning and motor execution, and how these
levels fit within the cognitive phonology approach. Ball et al. (2004), and Code
et al. (2006) using evidence from other language modalities, suggest that this
client displayed disruption to the motor execution level at first, but the progressive
nature of the disorder led to impairments at the level of stored organization as
well. It is worth noting that whereas all the client’s disorders appear to have the
same source (initially outside the phonology, but eventually within it), the effects
were sometimes phonetic (reduction of aspiration in fortis stops), and sometimes
phonological in the sense of loss of contrast (change of nasal to oral stops). This
reinforces the need to keep cause and effect separate.
11.9 Desiderata Revisited
Earlier we outlined desiderata for a descriptive clinical phonology and for a psy-
cholinguistic clinical phonology. We can now consider how far we can meet these
with the direction we have advocated here: a combination of gestural phonology
formalism within a cognitive approach to phonology. The first descriptive criterion
is theory independent; the requirement for working with naturalistic speech out-
put is vital irrespective of analytic framework. The second point was the need to
identify differences between target and realization. The gestural scores (as shown
in Figures 11.1 and 11.2) clearly show these differences, and are able to be com-
bined to show patterns affecting similar sounds, thus satisfying the third criterion.
The final descriptive desideratum (an evaluation metric) has not been met in what
we have discussed here, and is clearly an area for future research. We could envis-
age a metric based on amount of disruption to gestures or, perhaps more useful
in disordered child phonology, one based on developmental norms for gestural
coordination.
Turning to the modelling of speech behaviours, the first criterion refers to the
distinction between disruption to the phonology and disruption to the phonetics.
We have argued above that much of what is traditionally termed child phonologi-
cal disorder is in reality motor-based and that covert contrasts may well exist
between forms that sound like phonological mergers; and that gestural scores can
diagram this. As we noted earlier, Bybee’s (2001) account of cognitive phonology
does not prioritize the notion of contrast or make a distinction between processes
traditionally thought of as phonetic and those thought of as phonological. Clearly,
however, phonological disorders (such as paraphasias) do exist, and we need
further work to see how best to characterize these in the model we suggest here.
The second criterion refers to the difference between source and effect of error,
and as the gestural formalism is overtly concerned with speech production, the
articulatory setting should be clearly displayed (where this can be ascertained, of
course). Figures 11.1 and 11.2 show the different phonetic characteristics of target
/t/ and /k/, even though the effect was a phonological one. The surface effect is
captured via any traditional phonological account – whether it be rules, constraints
or processes. The third criterion concerns the provision of a psycholinguistically
adequate account, and we have argued that the combination of gestural phono-
logy with Bybee’s phonology in use provides this, just as it will guide remediation
(the fourth point) – though research is needed to ascertain how useful such an
approach would be.
11.10 Conclusion
Many of the traditional models of clinical phonology we have referred to are able
to meet the first set of desiderata we listed above: those concerned with the descrip-
tion of data. Few have been able to meet the second set: the modelling of speech
behaviours. The need for informed remediation means we should move towards
a way of distinguishing between cause and effect, and between patterns residing in
representation and in sound production (phonological and articulatory disorders).
We suggest that one profitable path for further research is via a cognitive phono-
logy coupled to a gestural description.
Notes
1 It could be argued that using a tripartite division, Dodd’s (1995) delayed group
would have a phonological disorder, her deviant consistent group could have a
phonetic planning disorder, and her deviant inconsistent group an articulatory dis-
order.
2 Lakoff (1993) and Wheeler and Touretsky (1993) describe a cognitive phonology
based within the traditions of Cognitive Grammar. This approach is clearly also
worthy of investigation by clinical linguistics, but Bybee’s (2001) model is the one
explored here.
3 The following sections are based on Ball (2003).
References
Anderson, J. and Durand, J. (1986), ‘Dependency phonology’, in J. Durand, (ed.),

Dependency and Non-Linear Phonology. London: Croom Helm, pp. 1–54.
Anderson, J. and Durand, J. (eds) (1987), Explorations in Dependency Phonology. Dordrecht:
Foris.
Anderson, J. and Ewen, C. (1987), Principles of Dependency Phonology. Cambridge:
Archangeli, D. (1988), ‘Aspects of underspecification theory’, Phonology, 5, 183–208.
Archangeli, D., and Langendoen, D. T. (eds), (1997), Optimality Theory: An Overview.
Oxford: Blackwell Publishers.
Baker, E., Croot, K., McLeod, S. and Paul, R. (2001), ‘Psycholinguistic models of speech
development and their application to clinical practice’, Journal of Speech Language
Hearing Research, 44, 685–702.
Ball, M. J. (2003), ‘Clinical applications of a cognitive phonology’, Phoniatrics, Logope-
dics, Vocology, 28, 63–69.
Ball, M. J., Code, C., Tree, J., Dawe, K. and Kay, J. (2004), ‘Phonetic and phonological
analysis of progressive speech degeneration: a case study’. Clinical Linguistics and
Phonetics, 18, 447–462.
Ball, M. J. and Müller, N. (2002), ‘The use of the terms phonetics and phonology
in the description of disordered speech’, Advances in Speech-Language Pathology, 4,
95–108.
Ball, M. J., Müller, N. and Damico, H. (2003), ‘Double onset syllable structure in a
speech disordered child’, Advances in Speech-Language Pathology, 5, 37–40.
Ball, M. J., Rutter, B. and Code, C. (2008), ‘Phonological analyses of a case of progres-
sive speech degeneration’, Asia-Pacific Journal of Speech, Language and Hearing,
11, 305–312.
Bauman-Waengler, J. (2004), Articulatory and Phonological Impairments: A Clinical Focus.
Boston: Pearson.
Bradford, A. and Dodd, B. (1996), ‘Do all speech-disordered children have motor
deficits?’ Clinical Linguistics and Phonetics, 10, 77–101.
Browman, C. P. and Goldstein, L. (1986), ‘Towards an articulatory phonology’, Phono-
logy Yearbook, 3, 219–252.
Browman, C. P. and Goldstein, L. (1992), ‘Articulatory phonology: an overview’,
Phonetica, 49, 155–180.
Bybee, J. (1985), Morphology: A Study of the Relation between Meaning and Form. Amsterdam:
John Benjamins.
Bybee, J. (2001), Phonology and Language Use. Cambridge: Cambridge University Press.
Chomsky, N. (1965), Aspects of a Theory of Syntax. Cambridge, MA: MIT Press.
Chomsky, N. (1972), Language and Mind. New York: Harcourt Brace Jovanovich.
Chomsky, N. and Halle, M. (1968), The Sound Pattern of English. New York: Harper and
Row.
Clements, G. N (1985), ‘The geometry of phonological features’, Phonology Yearbook, 2,
225–252.
Code, C. and Ball, M. J. (1988), ‘Apraxia of speech: the case for a cognitive phonetics’,
in M. J. Ball (ed.), Theoretical Linguistics and Disordered Language. London: Croom
Helm, pp. 152–67.
Code, C. F. S., Müller, N., Tree, J. T. and Ball, M. J. (2006), ‘Syntactic impairments
can emerge later: Progressive agrammatic aphasia and syntactic comprehension
impairment’, Aphasiology, 20, 1035–1058.
Dinnsen, D. and Barlow, J. (1998), ‘On the characterization of a chain shift in normal
and delayed phonological acquisition’, Journal of Child Language, 25, 61–94.
Dinnsen, D. and Gierut, J. (2008), ‘The predictive power of optimality theory for phono-
logical treatment’, Asia-Pacific Journal of Speech, Language and Hearing, 11, 239–249.
Dodd, B. (1995), The Differential Diagnosis and Treatment of Children with Speech Disorder.
London: Whurr.
Donegan, P. J. and Stampe, D. (1979), ‘The study of natural phonology’, in D. A. Dinnsen
(ed.), Current Approaches to Phonological Theory. Bloomington: Indiana University Press,
pp. 126–173.
Gibbon, F. (1990), ‘Lingual activity in two speech disordered children’s attempts to
produce velar and alveolar stop consonants: Evidence from electropalatographic
(EPG) data’, British Journal of Disorders of Communication, 25, 329–340.
Gibbon, F. E. (1999), ‘Undifferentiated lingual gestures in children with articulation /
phonological disorders’, Journal of Speech, Language, and Hearing Research, 42,
382–397.
Gibbon, F. E. (2007), ‘Research and practice in developmental speech disorders’,
in M. C. Pennington (ed.), Phonology in Context. Basingstoke: Palgrave Macmillan,
pp. 245–273.
Gibbon, F. E. and Scobbie, J. M. (1997), ‘Covert contrasts in children with phonolo-
gical disorder’, Australian Communication Quarterly, Autumn 1997, 13–16.
Gibbon, F. E. and Wood, S. E. (2001), ‘Undifferentiated gestures and articulatory drift
in the speech of children with articulation/phonological disorders’, Speech Motor
Control in Normal and Disordered Speech. Proceedings of the 4th International Speech Motor
Conference, pp. 52–56.
Gierut, J. A. (1990), ‘Differential learning of phonological oppositions’, Journal of
Speech and Hearing Research, 33, 540–549.
Gierut, J. A. (1999), ‘Syllable onsets: Clusters and adjuncts in acquisition’, Journal of
Speech, Language and Hearing Research, 42, 708–726.
Goldman, R. and Fristoe, M. (2000), Goldman-Fristoe Test of Articulation-2. Upper Saddle
River, NJ: Pearson.
Goldsmith, J. (1990), Autosegmental and Metrical Phonology. Oxford: Blackwell.
Grunwell, P. (1985a), ‘Comment on the terms “phonetics” and “phonology” as applied

in the investigation of speech disorders’, British Journal of Disorders of Communication,
20, 165–170.
Grunwell, P. (1985b), Phonological Analysis of Child Speech. Windsor: NFER-Nelson.
Grunwell, P. (1987), Clinical Phonology. 2nd ed. London: Chapman and Hall.
Grunwell, P. (1988), ‘Phonological assessment, evaluation and explanation of speech
disorders in children’, Clinical Linguistics and Phonetics, 2, 221–252.
Grunwell, P. (1992a), ‘Assessment of child phonology in the clinical context’, in
C. Ferguson, L. Menn and C. Stoel-Gammon, C. (eds), Phonological Development:
Models, research, implications. Timonium, MD: York Press, pp. 457–483.
Grunwell, P. (1992b), ‘Principled decision-making in the remediation of children with
developmental phonological disorders’, in P. Fletcher, and D. Hall (eds), Specific
Speech and Language Disorders in Children. London: Whurr, pp. 215–240.
Grunwell, P. (1997), ‘Natural Phonology’, in M. J. Ball, and R. D. Kent (eds), The
New Phonologies. San Diego: Singular, pp. 35–75.
Harris, J. (1990), ‘Segmental complexity and phonological government’, Phonology, 7,
255–300.
Harris, J. (1994), English Sound Structure. Oxford: Basil Blackwell.
Harris, J. and Cottam, P. (1985), ‘Phonetic features and phonological features in
speech assessment’, British Journal of Disorders in Communication, 20, 61–74.
Harris, J. and Lindsey, G. (1995), ‘The elements of phonological representation’, in
J. Durand and F. Katamba (eds), Frontiers of Phonology. London: Longmans,
pp. 34–79.
Harris, J., and Lindsey, G. (2000), ‘Vowel patterns in sound and mind’, in N. Burton-
Roberts, P. Carr and G. Docherty (eds), Phonological Knowledge, Conceptual and
Empirical Issues. Oxford: Oxford University Press, pp. 185–205.
Hawkins, P. (1985), ‘A tutorial comment of Harris and Cottam’, British Journal of
Disorders of Communication, 20, 75–80.
Hewlett, N. (1985), ‘Phonological versus phonetic disorders: Some suggested modifi-
cations to the current use of the distinction’, British Journal of Disorders of Communica-
tion, 20, 155–164.
Kamhi, A. (2006), ‘Treatment decisions for children with speech-sound disorders’.
Language, Speech, and Hearing Services in Schools, 37, 271–279.
Kaye, J., Lowenstamm, J. and Vergnaud, J.-R. (1985), ‘The internal structure of phono-
logical elements: A theory of charm and government’, Phonology Yearbook, 2,
305–328.
Kaye, J., Lowenstamm, J. and Vergnaud, J.-R. (1990), ‘Constituent structure and
government in phonology’, Phonology, 7, 193–232.
Kenstowicz, M. (1994), Phonology in Generative Grammar. Oxford: Blackwell.
Kent, R. D. (1996), ‘Developments in theoretical understanding of speech and its
disorders’, in M. J. Ball, and M. Duckworth, M. (eds), Advances in Clinical Phonetics.
Philadelphia: Benjamins, pp. 1–26.
Kiparsky, P. (1982a), ‘Lexical morphology and phonology’, in I.-S. Yang (ed.), Linguis-
tics in the Morning Calm. Seoul: Hanshin, pp. 3–91.
Kiparsky, P. (1982b), ‘From cyclic phonology to lexical phonology’, in H. van der Hulst,
and N. Smith (eds), The Structure of Phonological Representations (Part I). Dordrecht:
Foris, pp. 131–175.
Kiparsky, P. (1985), ‘Some consequences of lexical phonology’, Phonology Yearbook, 2,
83–138.
Khan, L. M., and Lewis, N. (2002), Khan-Lewis Phonological Analysis – Second Edition.
Circle Pines, MN: AGS.
Lakoff, G. (1993), ‘Cognitive phonology’, in Goldsmith, J. (ed.), The Last Phonological
Rule. Chicago: Chicago University Press, pp. 117–145.
McCarthy, J. (2002), A Thematic Guide to Optimality Theory. Cambridge: Cambridge
University Press.
Müller, N., Ball, M. J. and Rutter, B. (2008), ‘Systemic phonology applied to an idiosyn-
cratic case of /r/ disorder’, Asia-Pacific Journal of Speech, Language and Hearing, 11,
269–281.
Öhman, S. (1966), ‘Coarticulation in VCV utterances: Spectrographic measurements’,
Journal of the Acoustical Society of America, 39, 151–188.
Prince, A. and Smolensky, P. (1993), Optimality Theory: Constraint Interaction in Genera-
tive Grammar. Technical Report No.2, Rutgers University Center for Cognitive
Science. Cambridge, MA: MIT Press.
Skahan, S. M., Watson, M. and Lof, G. L. (2007), ‘Speech-language pathologists’ assess-
ment practices for children with suspected speech sound disorders: Results of a
national survey’. American Journal of Speech-Language Pathology, 16, 246–259.
Shriberg, L. D. and Kwiatkowski, J. (1982a), ‘Phonological disorders II: A conceptual
framework for management’, Journal of Speech and Hearing Disorders, 47, 242–256.
Shriberg, L. D. and Kwiatkowski, J. (1982b), ‘Phonological disorders III: A procedure
for assessing severity of involvement’, Journal of Speech and Hearing Disorders, 47,
256–270.
Sosa, A. V. and Bybee, J. (2008), ‘A cognitive approach to clinical phonology’, in M. J. Ball,
M. Perkins, N. Müller and S. Howard (eds), Handbook of Clinical Linguistics. Oxford:
Blackwell.
Spencer, A. (1996), Phonology. Oxford: Blackwell.
Stampe, D. (1979), A Dissertation on Natural Phonology. New York: Garland.
van der Hulst, H. G. (1995), ‘Radical CV Phonology: The categorical gesture’, in
J. Durand and F. Katamba (eds), Frontiers of Phonology. Essex: Longman, pp. 80–116.
van der Hulst, H. G. (1996), ‘Radical CV phonology: The segment-syllable connection’,
in J. Durand and B. Laks (eds), Current Trends in Phonology: Models and Methods, vol 1.
CNRS/ESRI Paris X, pp. 333–363.
Wheeler, D. and Touretsky, D. (1993), ‘A connectionist implementation of cognitive
phonology’, in Goldsmith, J. (ed.), The Last Phonological Rule. Chicago: Chicago
Yavaş, M. (2003), ‘Role of sonority in developing phonologies’, Journal of Multilingual
Communication Disorders, 1, 79–98.
CHAPTER
12
Sign Linguistics, Sign Language
Learning and Sign Bilingualism
Gary Morgan and Bencie Woll
12.1 Introduction
Sign languages have been created by Deaf people coming together in groups all
over the world. The study of how Deaf people communicate in signs, as well as the
linguistic and cognitive processes inherent in sign languages themselves can inform
many fields of research. In this chapter we review some areas which are particularly
relevant for the fields of applied linguistics, bilingualism and language acquisition.
We begin by describing how British Sign Language (BSL) was first documented.
Next we look at how linguistic analysis has revealed the extent to which sign lan-
guages have different levels of organization including a phonological structure.
We then move on to examine some aspects of the study of signed languages which
are informing mainstream research in bilingualism and first language acquisition.
In the final section we look at the wider social context in which BSL is used includ-
ing Education and second language learning.
12.2 First Descriptions: A Primitive Universal Gesture Language
Although descriptions of the use of signs by deaf people are found in the first
century Mishnah (compilation of Jewish law) and earlier, the first systematic obser-
vations of sign language use by deaf people in Britain date back to the seventeenth
century. John Bulwer’s two books, Chirologia (1644) and Philocophus: Or the Deafe
and Dumbe Man’s Friend (1648), and George Dalgarno’s Ars signorum, vulgo character
universalis philosophica et lingua (1661) while describing signing and discussing the
independence of signed language from spoken language, introduce a common
belief that signs and gestures are natural, and hence universal. This belief is found
in virtually every text on the subject published up to the middle of the twentieth
century.
What though you cannot express your minds in those verball contrivances of
man’s invention; yet you want not speeche; who have your whole body for a
tongue, having a language more naturall and significant, which is common to
224 Gary Morgan and Bencie Woll
you with us, to wit, gesture, the general and universall language of human
nature. (Bulwer, 1648)
The deaf man has no teacher at all and though necessity may put him upon . . .
using signs, yet those have no affinity to the language by which they that are
about him do converse among themselves. (Dalgarno, 1661)
Belief in the universality of sign language persisted throughout the nineteenth

century, but it began to be recognized that signing could change over time to
appear less universal. Watson, the headmaster of the school for the deaf in London
in the first quarter of the nineteenth century, and Stout, a psychologist, described
this process of language change:
The naturally deaf do not always stop here with this language of pantomime.
Where they are fortunate enough to meet with an attentive companion or two,
especially where two or more deaf persons happen to be brought up together, it
is astonishing what approaches they will make towards the construction of an
artificial language. (Watson, 1809)
What strikes him most . . . are at once signs by which he knows these objects and
knows them again; they become tokens of things. And whilst he silently elabo-
rates the signs . . . he develops for himself suitable signs to represent ideas . . .
and thus he makes himself a language . . . a way for thought is already broken
and with this thought as it now opens out the language cultivates and forms
itself further and further. (Stout, 1863)
Perhaps surprisingly, the development of twentieth century linguistics saw a rever-

sion to earlier views about sign language. This shift can be understood in the con-
text of the emphasis of linguistics on the primacy of spoken language, in contrast
to philologists who had often considered spoken language to be a degenerate form
of written language. The view of spoken language as the source for all other lan-
guage forms led to a peremptory dismissal of anything other than spoken language
as of interest:
The celebrated sign-languages of the American prairies . . . are only dialects

(so to speak) of the gesture-language . . . This ‘gesture-language’ is universal
not only because signs are ‘self-expressive’ (their meaning is self-evident) but
because the grammar is international. (Tylor, 1895)
Gesture languages have been observed among the lower-class Neapolitans,
among Trappist monks . . . among the Indians of our western plains . . . and
among groups of deaf-mutes. It seems certain that these gesture languages are
merely developments of ordinary gestures and that any and all complicated or
not immediately intelligible gestures are based on the conventions of ordinary
speech. (Bloomfield, 1933)
Sign Linguistics and Sign Bilingualism 225
12.3 Modern Approaches
In 1960, Stokoe published his pioneering work on American Sign Language (ASL),
working within a Bloomfieldian structuralist model. Stokoe recognized that sign
languages had an internal ‘phonological’ structure comparable to that of spoken
languages. Sign language researchers since 1960 have identified five features of
the sign languages of deaf communities throughout the world:
z They are complex natural human languages

z They are distinct from gesture
z There is no universal sign language
z They have their own grammars
z They do not represent spoken language on the hands
In the nearly 50 years since Stokoe, a substantial literature on many different

sign languages has been developed. The Hamburg bibliographic database (http://
www.sign-lang.uni-hamburg.de/bibweb) of sign language research lists thousands
of entries on nearly a hundred different sign languages.
Early modern research on sign languages emphasized the underlying structural
similarities of spoken and signed languages, but more recent research has moved
towards recognition that there are systematic typological differences. These arise
mainly from the interaction of language form with modality. Phonological and
morphological structures differ, since sign languages exhibit a relatively high
degree of systematic correspondence between form and meaning (iconicity or
visual motivation) in comparison to spoken languages. There are also consistent
grammatical features in which signed languages differ from spoken languages.
Sign languages exploit the use of space for grammatical purposes, preferring
3-dimensionality in syntax, while spoken languages prefer linearization and affixa-
tion. Other differences arise from the properties of the articulators (there are two
active articulators in sign languages – the hands) and the differing properties
of the visual and auditory perceptual systems. Observation of such differences has
led most recently to active consideration of the extent to which the contrasting
typological properties of spoken and signed languages indicate that linguistic the-
ory may need to take greater account of modality (Meier et al., 2002). A further
step in the evolution of our understanding of sign languages has been happening
in the past ten years through investigations into the extent deaf people combine
gesturing into their communication when they sign. It is probably the case that
there are slots in the discourse structure where signers can switch to gesturing
e.g. when they want to show role play of a character in a story. Further research is
needed to understand how these two ways of communicating in the visual modality
are integrated.
It has also been noted that there is greater typological variation among spoken
languages than among signed languages. There are a number of possible explana-
tions for the grammatical similarity among sign languages which still remain to be
researched fully. Sign languages are relatively young languages; indeed, the recent
studies of Nicaraguan Sign Language (Kegl et al., 1999) suggest that sign languages
can arise and develop spontaneously in deaf communities over three generations;
iconicity as an organizing factor in the lexicon may result in greater similarity at
the lexical level (Woll, 1984); the linear syntax found in spoken languages may
intrinsically allow greater differences than spatial syntax; and the relatively low
percentage of signers who are themselves the children of signers results in contin-
ual recreolization with resulting similarity of grammar (Fischer, 1978). There is
evidence to support all of these hypotheses, but a great deal of research remains to
be done in this area.
12.4 Cross-Linguistic Communication: Studies of International Sign
Despite the underlying structural similarities among sign languages, unrelated

sign languages are mutually unintelligible. However, it has long been noted that
signers are able to communicate effectively across language boundaries (Figure
12.1). This observation is first recorded in an account of a visit to the deaf school
in London by Laurent Clerc, a deaf Frenchman and teacher of the deaf, who sub-
sequently became the first teacher of the deaf in the United States.
As soon as Clerc beheld this sight [the children at dinner] his face became ani-
mated; he was as agitated as a traveller of sensibility would be on meeting all
of a sudden in distant regions, a colony of his own countrymen . . . Clerc
Figure 12.1 Early twentieth century pamphlet illustration of cross-linguistic communica-

tion (Maginn, F. (n.d.) Guide to the Silent Language of the Deaf. Belfast: Francis Maginn).
approached them. He made signs and they answered him by signs. This unex-
pected communication caused a most delicious sensation in them and for
us was a scene of expression and sensibility that gave us the most heartfelt
satisfaction. (De Ladebat, 1815)
There have been a small number of studies of International Sign (IS), as it has
become known (Allsop et al., 1995; Supalla and Webb, 1995; Rosenstock, 2008).
These indicate that IS is a mixture of gesture and sign language, with a limited
lexicon and extensive paraphrasing. IS also has only a limited constituent struc-
ture, with no clear evidence for recursiveness, and it is difficult to identify sentence
boundaries. On the other hand, IS is not simply gesture. IS uses the same restricted
space which is found in sign languages; all of the paraphrases and circumlocutions
are formally possible signs in the signer’s own language; and the more experi-
enced an individual is with IS, the more extensive the paraphrasing, suggesting
increased awareness of the likely unintelligibility of structures taken from a spe-
cific sign language.
12.5 Bimodal Bilingualism
Until recently, all studies of bilingualism were studies of individuals and communi-
ties in which two spoken languages are used. With the development of research
on sign languages, it has become clear that bilingualism can exist in two forms:
bimodal (sometimes called ‘cross-modal’) bilingualism or unimodal bilingualism.
Unimodal bilingualism occurs when either two spoken or two sign languages are
used (e.g. Irish and British Sign Language); Cross-modal or bimodal bilingualism
occurs when the two languages are perceived by different sensory systems and
there are two different output channels; they thus exist in different modalities: one
signed and one spoken. Recognition of bimodal bilingualism leads to a necessary
re-evaluation of models of bilingualism and the following questions:
z What are the ramifications of removing a major articulatory constraint on bilin-

gual language production?
z Is code-switching the consequence of articulatory constraints?
Bimodal bilingualism differs from uni-modal bilingualism with respect to

the temporal sequencing of languages. Bimodal bilinguals produce code-blends,
rather than code-switches. They do not stop speaking to sign or stop signing to
speak. They produce sign and speech simultaneously when in a bilingual mode of
communication. In general, code-blends are semantically equivalent, suggesting
that code-blending is not produced in an effort to distribute distinct information
across modalities. Code-blends also differ from code-switches with respect to gram-
matical category.
Since in unimodal spoken language bilingualism, just one production system
(the vocal tract) is utilized, there is a ‘bottleneck’ in processing, which has been
identified both in behavioural and neuroimaging studies. For production, this is
evident in constraints on serial output: only one utterance can be produced at

any one time: selection between two competing languages with respect to the final
common output path may impose a greater cognitive burden than selection of
items within one language. For perception, similar constraints could apply – the
speaker’s utterances are usually processed by the ear, and require serial analysis to
be understood. For multilingual speech, a single (auditory) processing stream may
carry information in more than one language, imposing a selection burden on
specific cortical systems that may be somewhat independent of the languages
themselves, and more involved in selection between specific percepts/actions.
This is where people who are bilingual in a spoken and a signed language come
into the picture. For such bimodal bilinguals, vocal tract (speech) and manual-
gestural actions (signed languages) could – in principle – be processed simulta-
neously, as they use separate sets of articulators which are seen to be spatially
separated. That is, the serial order constraint may not apply in the same way to
bimodal as to unimodal bilinguals. The processing of speech and sign, within a
single proficient user of both languages, can therefore offer a unique insight into
core language processes. Few studies have explored the implications of this situa-
tion, yet bimodal bilingualism is common. Hearing children of Deaf parents
(sometimes called CODAs – children of deaf adults) can acquire a sign language
as a first language. As adults, they have full access to at least two languages: a visual-
manual one (signed language) and an auditory-vocal one (spoken language).
Among the few studies to have explored how bimodal bilingualism works in CODAs
are those of Emmorey et al. (2005). They have shown code blending – especially
of action terms – in the production of words and signs where these reflect a com-
mon conceptual source. Code blending reflects the simultaneous use of sign and
word in a single utterance. This is, of course, not possible for unimodal bilinguals,
who must sequence the language items in production. Bimodal bilinguals can
and do escape the serial order constraint in production. However, they also tailor
these outputs to the communicative situation. Bimodal blends were produced
spontaneously in CoDA–CoDA interactions, but not when CoDAs communicated
with monolinguals.
More numerous than CoDAs, however, are deaf people who acquire a signed
language at home or at school, but also use a spoken language, relying on
speechreading, hearing aids and voice training. Although it is more onerous for
the deaf person to master a spoken than a signed language, speechreading affords
a window into speech which, while sometimes obscure, is nevertheless useful. What
is more, there are powerful social pressures to develop some degree of speech
mastery, so that among deaf people who have acquired a sign language, bimodal
bilingualism is the norm in those countries where deaf children receive education.
Deaf people never live in communities which do not use speech, and they must
acquire some degree of competence in understanding speech and in making
their speech understood. Mastery of both the signed language of the Deaf com-
munity and the spoken/written language of the hearing community is the goal of
deaf bilingual education, since the bedrock of education is literacy.
It is perhaps strange, therefore, that neurolinguistic studies of the cortical organi-
zation of sign language have not taken into account that the deaf communities of
Europe and North America are bilingual communities, and that the great majority
of deaf people studied are, to a greater or lesser degree, bimodal bilinguals. The
same point can be made in studies of individuals known as Homesigners (deaf
children who do not have sign language input at home and create gestures to
communicate with their parents). It is sometimes assumed that these children
are totally devoid of language input. Despite spoken language input being
impoverished for these children they are still exposed to some language. To date,
most studies have compared sign language processing (in deaf people or hearing
native signers) to spoken language processing (in hearing monolinguals). The
contrasts between signed and spoken language processing with these groups
suggest very similar brain regions are implicated in each. However, these studies
have all adopted a monolingual perspective, considering deaf sign language users
to be equivalent in terms of their language experience to hearing monolinguals.
Research on bimodal bilinguals with aphasia are also likely to prove of great
interest. There is one recorded case study of aphasia in a bimodal bilingual patient
(Marshall et al., 2005). ‘Maureen’, a Deaf adult, suffered a left middle cerebral
artery infarct at the age of 70, and showed profound impairments in semantic
processing both of BSL signs and English words. In addition to her dense impair-
ment in comprehending signs or speech, she was able to utter occasional English
words but produced no BSL signs. Nevertheless, observation of BSL signs could
sometimes cue the production of English words, suggesting a level of lexical ‘cross-
talk’ between the languages, and a different relationship between the lexicons
of the bimodal bilingual’s two languages, than has been reported for unimodal
bilinguals. This may be related to the fact that it is possible to sign and speak at the
same time in simultaneous communication, which is not possible between two
spoken languages. Further studies are needed to explore this topic. Those studies
will need to use more demanding linguistic tasks, requiring greater levels of lan-
guage selection, in order to discern the extent to which the serial order constraint
may apply in bimodal bilingual language processing. By contrast with unimodal
bilingual language processing, bimodal bilingual processing may need to allocate
special attention effectively to distinguish manual signs and speech. Again, the use
of more demanding linguistic tasks, including translating between languages, or
distinguishing a target delivered by speech and signs when both are seen simulta-
neously, might indicate whether special resources are called upon to manage this
unique requirement, and may identify the cortical networks that support them. In
addition, the study of SL unimodal bilingualism, i.e., of people who have access to
more than one SL, will offer further clarification of core and modality-specific
issues in bilingual processing.
12.6 Sign Language Acquisition
12.6.1 Acquisition of a Sign Language as a First Language
There are two motivations for studying sign language acquisition in children who
are exposed to rich and consistent sign language from birth. The first stems from
a scientific interest in language development and modality. The second is a more

practical reason. Since the vast majority of deaf children are born to hearing par-
ents, their exposure to language is very different from that of hearing children
learning a spoken language from hearing parents. Their hearing parents may
begin to learn a sign language when deafness is diagnosed, but their input may
be limited because of their poor sign language skills; if they choose to communi-
cate only through spoken language, the child may have only limited access to the
linguistic signal. Thus, the typical experience for the deaf child is late and impov-
erished exposure to a first language (Spencer, 1993; Harris, 2001). In order to
understand the impact of atypical input on language development, this section
focuses mainly on sign language acquisition in the minority of deaf and hearing
children who grow up with deaf parents and therefore have good sign language
models from birth.
Study of children’s acquisition of BSL has been part of the growing research
into Deafness and signed language use carried out over the past three decades.
During this time research has shown that BSL should be recognized as a natural
and complete language in terms of linguistic structure (e.g. Sutton-Spence and
Woll, 1999), brain processing (e.g. MacSweeney et al., 2002) and acquisition
processes (Morgan and Woll, 2002). Although more research is needed about
the specific features of the early stages of BSL acquisition, there has been general
agreement that sign language acquisition parallels that of spoken language
(Newport and Meier, 1985; Schick, 2003; Mayberry and Squires, 2006). There is
also agreement that the developmental course from first words and word combina-
tions to sentences is similar across sign languages (Mayberry and Squires, 2006)
although the linguistic details vary across sign languages. Young deaf or hearing
children exposed to a sign language by their deaf parents become proficient
signers and the developmental stages are comparable to hearing children exposed
to a spoken language by their hearing parents (Morgan and Woll, 2002). Some
longitudinal observational studies have revealed specific interactional features in
this process, e.g., the importance of eye contact and turn-taking (Harris and Mohay,
1997) and there have been several studies of characteristics of infant-directed sign-
ing, such as increased size of sign movement and repetition (Masataka, 1996).
Similar adaptations are also found in child-adjusted spoken language.
12.6.2 Input
When fluent adult signers interact with young children they often adapt their
signing to suit the needs of their young conversation partner. Many, but not all, of
the characteristic features of child directed signing are the same as those observed
in child directed speech or motherese. Adult–child signing is slower, more exag-
gerated with larger movement and tends to have more repetition than adult–adult
interaction (Masataka, 1996). One of the major differences between signed and
spoken language is that signers encode grammatical distinctions by the physical
movement of a sign through sign space rather than the production of a spoken
word with affixes. This means that children are required to pay attention visually
to adult’s signing and conversely adults have to time their signing to moments
when children are looking at them. These two constraints, (1) looking at the com-
munication partner and (2) timing the input to the child at moments when the
child is looking, are not necessary in spoken language acquisition. Hearing chil-
dren can pick up spoken language without having to pay close attention all of the
time to adults who are speaking to them. These constraints have consequences for
certain modality specific characteristics of child directed signing. (Harris, 2001).
One common finding from several studies of early child-parent sign language
interaction is that the number of signs addressed to young children by fluent
signing adults using ASL, BSL and Sign Language of the Netherlands (SLN) is
considerably less than that in hearing parent–hearing child interaction in the same
community and also less than in adult–adult sign interaction (e.g. Kantor, 1982;
Kyle and Ackerman, 1990; Gallaway and Woll, 1994; van den Bogaerde and Baker,
1996; Harris, 2001). Although deaf children receive considerably less language
input from fluent adult signers there is no evidence that this slows language devel-
opment down in signing children. As stated previously, signing and speaking
children reach the major language development milestones at the same ages.
Child directed signing contains fewer signs because it must be linked to those
moments that the child’s gaze is at the adult signer, and it may contain child
directed modifications to the grammar such as the dropping of agreement or
aspectual inflections on verbs. Adult signers attract the visual attention of children
(usually by waving or touching the child) before they sign. This difference in the
amount of input directed to children who are learning sign compared with hear-
ing children’s constant exposure to ambient speech raises the question as to how
much language is necessary for successful, full development of language.
12.6.3 First Signs – Iconicity in First and Second Language Development
Many BSL signs exhibit an arbitrary mapping between their form and meaning
(e.g. MOUSE in BSL – where a bent index finger makes contact and twists at the
side of the nose) however the majority of signs are iconic (e.g. TELEPHONE in BSL –
which resembles the shape of a telephone as the fist has thumb and little finger pro-
truding) (Sutton-Spence and Woll, 1999). This direct link between meaning and
form may influence children’s development of sign language as first languages.
Before discussing the data on first signs it is important to explore a little the
notion of iconicity. First we need to ask the question: ‘iconic’ for whom? Iconicity
is often culturally determined (Pizzuto and Volterra, 2000). The sign for BREAD
in BSL makes reference to the cutting of a hand-size slice while in French Sign
Language the same concept is encoded by the action of putting something
(a baguette) under your arm. Even clearly iconic signs often make reference to
only part of an object’s visual appearance (e.g. BIRD in BSL makes reference to
a beak) so further abstraction may be required to see a relationship.
There are also differences in how much iconicity is accessible to sign learners
depending on whether they are an adult or a young child. Are 12–24-month olds as
equally able to ‘see’ iconic links as adults who are learning sign as a second language?
It is plausible that adult learners will use available world knowledge based on some
iconic property in learning the meaning of a sign or in guessing the meaning for
a novel sign. For example, the sign for Paris in BSL refers to the shape of the Eiffel
tower. Information of this kind is not available to the young child because of
restricted world knowledge.
Studies of children acquiring ASL have reported that less than a third of chil-
dren’s first signs are closely linked to iconic properties of the referent. Children’s
first signs are more likely to be associated with the same sets of semantic categories
evident in children’s early speech e.g. signs for people, animals and food (Anderson
and Reilly, 2002). In a study of first language acquisition of BSL from age 1;10 – 3;0,
Morgan, Barriere and Woll (2003) reported that from age 1;10 onwards the majo-
rity of signs that were being acquired by the child were non-iconic. Children who
acquire a sign language as their first language therefore seem blissfully unaware of
the ‘etymology’ of a sign such as MILK (which imitates the action of milking a
cow). It is probably only later (say after age 3 years) once the core grammar has
been mastered, that young children might return to the language forms they have
learned previously and carry out a metalinguistic analysis akin to what happens
when children discover overtly the segmentation properties of their language dur-
ing the development of literacy (Morgan, 2005).
Early research on ASL from Bonvillian et al. (1993), Prinz and Prinz (1979) and
others, reported native signing deaf and hearing children’s first signs appeared
1–3 months prior to the first words of hearing peers, at around 7–9 months of
age or earlier. Other research reported that first signs emerge closer to the magic
12 months and at the same age as hearing children’s first words (Marschark, 1993:
Chapter 5, for reviews). Later research on ASL, LIS and BSL (Meier and Newport,
1990; Petitto, 1992; Volterra and Caselli, 1985; Woll, 1998) that distinguishes
between prelinguistic gesture, babbling and the first symbolic and intentionally
communicative use of signs has not replicated the findings from the early studies.
Across different studies, first signs have been reported from as early as 8 to as late
as 16 months. Children reach the 10 sign stage at around 12 months and then go to
learn their first 50 signs from 24 months and older (Anderson and Reilly, 2002).
It is perhaps easy to see how deaf parents when reporting on their deaf child’s
signing might see prelinguistic gesture as possible signs. One frequently reported
early sign in ASL exposed children, is MILK (made by the opening and closing
of the 5 hand in ASL). However open-close has previously been reported as one
of the first hand-internal movements to appear in manual babbling in ASL (e.g.
Cheek et al., 2001), hearing children’s gestures (Volterra et al., 2005) and chil-
dren’s first signs (Cheek et al., 2001). Marschark (2002) suggests that deaf parents
may pay more direct attention to their children’s manual productions than is nor-
mally the case for hearing child–adult interaction thus ‘spotting’ gestures that fit a
particular communicative function earlier. Further as Meier and Newport (1990)
suggested, the fact that all of a sign’s component parts are visible to parents and
researchers compared with significantly less of children’s first spoken words may
lead some to see partial signs earlier.
Importantly when Meier and Newport (1990) reviewed the studies of early sign-
ing they concluded that any perceived advantages disappear by the two-sign stage.
So if we go beyond word learning as an indicator of language development and
consider children’s acquisition of syntax and morphology and the interaction
between morpho-syntax and semantics, there is no evidence so far that children
exposed to a sign language from infancy find the acquisition of these abstract prin-
ciples any easier than children acquiring a spoken language (see also Meier, 1987).
The difficulty in deciding if hand movements of young children are first signs or
gesturing may explain some of the claims made in the media about the develop-
mental advantage of ‘baby signs’ with hearing children.
12.6.4 Baby Signing
In recent years, the teaching of signs to hearing infants has become increasingly
popular. Baby signing courses are in the main taught to hearing parents of hearing
babies and involve speaking and signing at the same time. Hearing teachers nor-
mally use a repertoire of signs (sometimes from foreign sign languages such as ASL
rather than BSL) and integrate the signs into songs and nursery rhymes. There are
many extravagant claims (made by the producers of Baby Sign materials) that Baby
signing courses will lead to higher levels of IQ, literacy and even reduce toddler
tantrums; however, there has been very little systematic research to back these
claims up (see Johnston et al., 2005 for a review).
We know that babies develop language (signed or spoken) in situations
where carers offer sensitive contingent responses, and appropriate stimulation for
requests, naming, communication of feelings and social behaviours. The argument
for the introduction of sign to babies rests on research suggesting that babies
learning a sign language as a first language produce their first signs slightly earlier
than babies produce their first spoken words. Several explanations have been put
forward for this reported sign advantage. One suggestion is the differing matura-
tion rates of the motor system for hand movements when compared with that of
the speech apparatus (Bonvillian and Siedlecki, 1996). However, these early signs
actually seem formationally identical to the gestures produced by both deaf and
hearing babies – hand opening and closing for more or milk for example. So the
‘signs’ are actually gestures that all babies produce. Acredolo and Goodwyn them-
selves state on their website: ‘Research has shown that signs are easiest for babies –
and for parents – when they involve simple gestures and when they resemble the
things they stand for (e.g. fingers to lips for eat, arms out straight like wings for air-
plane)’ http://www.babysigns.com.sg/FAQ.html.
If specific gestures and sounds are reinforced, babies will develop communica-
tion skills in both modalities. It does seem to be the case that there is a short period
during which infants can produce more differentiated and controlled hand ges-
tures than speech sounds – but it is a transitory phase. By the time children have
started to produce their first words and then combine them, the timelines for signs
and words appear to be essentially similar.
12.6.5 BSL and Education
The use of BSL in education in the United Kingdom is subject to wide variation
and application. Since the 1978 Warnock report into Special Educational Needs
there has been a very strong move towards mainstreaming deaf children; although
at present it is not clear if this is an appropriate system for all deaf children to fol-
low. Some deaf children informally learn sign language once they arrive at primary
schools but are neither formally taught the language, nor are exposed to it as a
language of instruction. Their access to BSL may be via school staff and pupils who
vary greatly in their competence. Currently Communication Support Workers,
who in some cases provide the deaf child with his only sign language input have
very low levels of signing fluency. A minority of deaf children go to schools with
bilingual education policies; out of 25,020 deaf children in school in the United
Kingdom in 2000, 109 attended schools that considered themselves monolingual/
bimodal, and 26 attended schools that considered themselves bilingual/bimodal
(BATOD, 2000). Without distinguishing children by hearing levels and also the
quality of various schools, 19,753 children (around 80 per cent) attended mono-
lingual oral schools while 1,114 (4.45 per cent) attended schools with a sign bilin-
gual policy that was BSL dominant. This is in stark comparison with Scandinavian
education, where almost all deaf children receive bilingual education. From the
age of diagnosis, their families are given sign language classes with support in mak-
ing choices about their deaf children. There are also government funded family
centres that teach their children sign language at reception age (Mahshie, 1995).
There are few standardized assessment tools for BSL acquisition. Herman and
Roy (2006) report on language achievement scores using the BSL Receptive Skills
test (Herman, Holmes and Woll, 1999), which assesses comprehension of BSL
morphosyntax. They found that many deaf children do not achieve age-appropriate
levels of BSL. Using normative data from 135 children, the authors found that in
a wider sample of 181 deaf children (age range 40–177 months, mean 102.97 months),
deaf children from hearing families (n = 113) had lower BSL comprehension
scores compared with their native signing peers (n = 35, p < .001). These results
are similar to those found in other studies that consistently show superior sign
language skills in deaf children from deaf families (Paul and Quigley, 2000), as
well as persistent inadequacies in the language environment provided by educa-
tion systems that report using sign language (Greenberg and Kusché, 1987; Mead-
owet al., 1981; Meadow, 1980; Barlow and Brill, 1975). Deaf children from hearing
families experience great variability in exposure to BSL in education, including
a variety of approaches to communication and often limited access to deaf role
models. Singleton and Morgan (2005) observe that deaf native signers arrive at
school ‘knowing how to learn’, in contrast to late learners of sign language who
may begin school with limited language. In some schools teachers actively attempt
to develop deaf children’s metalinguistic awareness of the differences and similari-

ties between BSL and English (Morgan, 2005).
12.6.6 British Sign Language and the Outside World:

Learning BSL as an L2
Over the past 25 years there have been substantial changes in the relationship of
BSL to the wider public. Among the most striking changes have been the represen-
tation of BSL in the media. As recently as 1989, the following review ridiculing the
sign language interpreted version of the Queen’s Christmas message appeared in
the Guardian:
The funniest thing I saw was the Queen’s Speech interpreted into sign-language
by a splendid . . . lady. It was an Oscar winner among sign-language mimes . . .
I hope HM and millions of the deaf enjoyed it as much as we did. But my guess
is some back-room electronic wizard is making urgent inquiries about emigration.
(The Guardian, 2 January 1989)
The ridiculing of the community and its language was also found among educa-
tors of deaf children until relatively recently:
I had a lot of punishments for signing in classrooms and at playground . . . then

one morning at assembly I was caught again, then, ordered to stand in front.
The headmistress announced that I looked like a monkey . . ., waving my hands
everywhere. She [said] she will put me in a cage in the zoo so the people will
laugh at a stupid boy in the cage. (Kyle and Woll, 1985: 263)
Despite the limited acceptance of BSL by educators, there has been an enormous
increase in the numbers of hearing people learning BSL in recent years. This can
be seen from Figure 12.2 which shows the numbers of students taking national
examinations offered by the Council for the Advancement of Communication with
Deaf People at Stage 1 and 2 from 1995–2006. The cumulative total of people with
at least a basic knowledge of BSL are now over 190,000 (a significantly larger
number than the total of the Deaf community itself). There are now plans to offer
a BSL GCSE, which is also likely to prove popular.
12.6.7 BSL-English Interpreters
The increase in numbers of students taking BSL courses has not been matched by
a corresponding increase in the number of BSL-English interpreters. Indeed, the
shortage of interpreters is one of the most serious problems facing Deaf people,
since interpreters enable access to communication with the hearing world. For
example, the Digital Broadcasting Act requires the provision of BSL on 10 per cent
of all digital terrestrial programming; the Disabled Students Allowance provides
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
1995 1996 1997 1999
2000 2001 2002
2003 2004 2005
BSLStage 2 BSLStage 1
Figure 12.2 Students taking BSL exams (Woll, 2001. Additional data provided by Signa-
ture: the Council for the Advancement of Communication with Deaf People).
funding for sign language interpretation for undergraduate and postgraduate stu-
dents, and the Disability Discrimination Act requires the provision of sign lan-
guage interpretation by firms and government for publicly available services.
However, as can be seen from Figure 12.3 below, there has been a relatively slow
increase in the number of qualified sign language interpreters over the past 24
years. The recent rapid rise in the number of qualified interpreters reflects short-
term projects that have sought to move trainees to qualified status. However, the
number of trainees has not risen, suggesting that the supply of interpreters will not
increase substantially in the next few years.
12.6.8 Official Recognition of BSL
The British Deaf community campaigned for many years for official recognition of
BSL by the UK government which finally happened in 2003. This recognition
brought little tangible change, although the Disability Discrimination Act has had
substantial impact in providing funding for interpreting. With the signing
by the government of the European Charter for Regional or Minority Languages,
the campaign has moved towards seeking the inclusion of BSL on the Charter
list of minority languages, in order to ensure adequate funding for training and
350
300
250
200
150
100
50
0
1982 1987
1992 1995
1997 2000
2003 2006
Qualified Interpreters Trainee Interpreters
Figure 12.3 Numbers of qualified and trainee interpreters (Data provided by Signature: the
Council for the Advancement of Communication with Deaf People).
provision of interpreters and acceptance of BSL in public settings such as the law
and education.
12.7 Conclusions
The history of BSL, like that of many minority languages, cannot be separated
from the study of its relationship with the majority language community which
surrounds it. At the beginning of the twenty-first century, there are two contrasting
futures: on the one hand, there are pressures, such as the decrease in opportuni-
ties for Deaf children to use BSL with their peers as a result of the move to main-
stream education, and the possible decrease in the Deaf population as a result of
medical intervention and advances in genetics; on the other hand, increased inter-
est and demand from the hearing community for courses in BSL, increased use of
BSL in public contexts such as television, baby sign courses and increased pride of
the Deaf community in their distinctive language and culture. It is to be hoped
that BSL will continue to be a living language.
Where next in the study of Deaf people and signed languages? The past ten years
have seen the growth in our understanding of the neurological pathways involved
in processing language versus other complex information systems. Technological
advances in analysis (e.g., motion capture and eye-tracking) have enabled an
improved understanding of sign language structure and the online processing of
the linguistic components of signing. Looking at how sign language is perceived
and processed in imaging studies has made a substantial contribution to illuminat-
ing the relationship of language form to communication channel. In the next ten
years we will see more sophisticated uses of brain imaging techniques to explore
brain plasticity, particularly in situations of late first and second language acquisi-
tion, and in contrasting and comparing bimodal and unimodal bilingualism. The
growth of interest in gestural communication and in the evolution of human lan-
guage have both contributed to and been enriched by sign language research, and
these fields are likely to draw more closely together in the future. At a more applied
level a review and re-examination of the educational experiences and achieve-
ments of Deaf children is overdue. Despite the advent of universal mainstreaming
of all children with disabilities, and new technologies such as cochlear implants,
there appears to have been little improvement in the literacy levels of Deaf school-
leavers. New research into language development and the learning of literacy and
numeracy skills needs to feed into this educational policy review.
References
Allsop, L., Woll, B. and Brauti, J. M. (1995), ‘International Sign: The creation of an
international deaf community and sign language’, in H Bos and T Schermer (eds),
Sign Language Research 1994, Hamburg: Signum Press, pp. 171–188.
Anderson, D. and Reilly, J. (2002), ‘The MacArthur Communicative Development
Inventory: Normative data for American Sign Language’, Journal of Deaf Studies and
Deaf Education, 7, 2, pp. 83–119.
Barlow, I. H. and Brill, R. G. (1975), ‘An evaluation of reading and academic achieve-
ment levels of 16 graduating classes of the California School for the Deaf, River-
side’, Volta Review, 77, 255–266.
The BATOD Survey: Summary of data from England 2000. http://www.batod.org.uk/
index.php?id=/publications/survey/2000-summary.pdf, accessed 15.05.08
Bloomfield, L. (1933), Language. New York: Holt Rinehart Winston.
Bonvillian, J. and Siedlecki, T. (1996), ‘Young children’s acquisition of the location
aspect of ASL signs: parental report findings’, Journal of Communication Disorders, 29,
13–25.
Bonvillian, J., Orlansky, M. and Novack, L. (1993), ‘Developmental milestones: Sign
language acquisition and motor development, Child Development, 54, 1435–1445.
Bulwer, J. B. (1644), Chirologia, or the Natural Language of the Hand. London: R Whitaker.
Bulwer, J. B. (1648), Philocophus: Or the Deafe and Dumbe Man’s Friend. London: Humphrey
Moseley.
Cheek, A., Cormier, K., Repp, A. and Meier, R.P. (2001), ‘Prelinguistic gesture predicts
mastery and error in the production of early signs’, Language 77, (2), 292–323.
Dalgarno, G. (1661), Ars signorum, vulgo characger universalis philosophica et lingua.
London: E. Hayes.
Emmorey, K., Borinstein, H. B. and Thompson, R. (2005), ‘Bimodal bilingualism:
Code blending between ASL and spoken English, in J. Cohen, K. T. McAlister,
K. Rolstad and J. MacSwan (eds), Proceedings of the 4th International Symposium on
Bilingualism. Somerville, MA: Cascadilla Press, 663–673.
Fischer, S. D. (1978), ‘Sign language and creoles’, in P. Siple (ed.), Understanding Language
through Sign Language Research. New York: Academic Press.
Gallaway, C. and Woll, B. (1994), Interaction and childhood deafness. In C. Gallaway

& B. Richards. (eds), Input and Interaction in Language Acquisition. Cambridge,
New York : Cambridge Univ. Pr. pp. 197–218.
Greenberg, M. and Kusché, C. (1987), ‘Cognitive, personal, and social development of
deaf children and adolescents’, in M. C. Wang, M. C. Reynolds and H. J. Walberg
(eds), Handbook of Special Education: Research and Practice: Volume 3. Low Incidence
Conditions. New York: Pergamon, 95–129.
Harris, M. (2001), ‘It’s all a matter of timing: sign visibility and sign reference in deaf
and hearing mothers of 18-month-old children’, Journal of Deaf Studies and Deaf
Education, 6, (3), 177–185.
Harris, M. and Mohay, H. (1997), ‘Learning to look in the right place: A comparison
of attentional behavior in deaf children with deaf and hearing mothers’, Journal of
Deaf Studies and Deaf Education, 2, 95–103.
Herman, R., Holmes, S. and Woll, B. (1999), Assessing British Sign Language Development:
Receptive Skills Test. Coleford: Forest Book Services.
Herman, R. and Roy, P. (2006), ‘Evidence from the extended use of the BSL receptive
skills test’, Deafness & Education International, 8, (1), 33–47.
Herman, R., Grove, N., Holmes, S., Morgan, G., Sutherland, H. and Woll, B. (2004), Assessing
BSL Development: Production Test (Narrative Skills). London: City University Publication.
Johnston, J., Durieux-Smith, A. and Bloom, K. (2005), ‘Teaching gestural signs to
infants to advance child development: A review of the evidence’, First Language, 25,
235–251.
Kantor, R. (1982), ‘Communicative interactions: Mother modification and child acqui-
sition of American Sign Language’, Sign Language Studies, 11, (36), 233–282.
Kegl, J., Senghas, A., and Coppola, M. (1999), ‘Creation through contact: sign lan-
guage emergence and sign language change in Nicaragua’, in M deGraff (ed.), Lan-
guage Creation and Language Change. Cambridge, MA: The MIT Press, pp. 179–237.
Kyle, J. and Ackerman, J. (1990), ‘Signing for infants: Deaf mothers using BSL in
the early stages of development’, in Edmondson, W. and Karlsson, F. (eds), Interna-
tional Studies on Sign Language and Communication of the Deaf 10. Hamburg: Signum,
pp. 200–211.
Kyle, J. G. and Woll, B. (1985), Sign Language: The Study of Deaf People and Their
Language. Cambridge: Cambridge University Press, p. 263.
MacSweeney, M., Woll, B., Campbell, R., McGuire, P. K., David, A. S., Williams,
S. C. R., Suckling, J., Calvert, G. A. and Brammer, M. J. (2002), Neural Systems Under-
lying British Sign Language and Audio-Visual English Processing in Native Users. Brain
125, 1583–1593.
Mahshie, S. (1995), Educating Deaf Children Bilingually. Washington, DC: Gallaudet Uni-
versity The Clerc Center.
Marschark, M. (1993), Psychological Development of Deaf Children. Oxford University Press.
Marschark, M. (2002), ‘Foundations of communication and the emergence of language
in deaf children’, G. Morgan and B. Woll (eds), Directions in Sign Language Acquisition.
Amsterdam: John Benjamins, 1–28.
Marshall, J., Atkinson, J., Woll, B. and Thacker, A. (2005), Aphasia in a Bilingual User
of British Signlanguage and English: Effects of Cross-linguistic Cues. Cognitive Neuropsycho-
logy, 22, 719–736.
Masataka, N. (1996), ‘Perception of motherese in a signed language by 6-month-old

deaf infants’, Developmental Psychology, 32, 874–879.
Mayberry, R. I. and Squires, B. (2006), ‘Sign language acquisition’, in E. Lieven (ed.),
Language. Acquisition, Vol. 11, Encyclopedia of Language and Linguistics, 2nd Edition.
Oxford: Elsevier, pp. 291–296.
Meadow, K. P. (1980), Deafness and Child Development. Berkeley: University of California
Press.
Meadow, K., Greenberg, M., Erting, C. and Carmichael, H.(1981), ‘Interactions of deaf
mothers and deaf preschool children’, American Annals of the Deaf, 126, 454–468.
Meier, R. P. (1987), ‘Elicited imitation of verb agreement in American Sign Language:
iconically or morphologically determined?’ Journal of Memory and Language 26. (3),
362–376.
Meier, R. P., Cormier, K. A. and Quinto, D. G. (eds) (2002), Modality and Structure in
Signed and Spoken Languages. Cambridge: Cambridge University Press.
Meier, R. P. and Newport, E. (1990), ‘Out of the hands of babes: On a possible sign
advantage in language acquisition’, Language, 66, (1), 1–23.
Morgan, G. (2005), ‘The development of narrative skills in British sign language’, in
B. Schick, M. Marschark and P. Spencer (eds), Advances in the Sign-Language
Development of Deaf and Hard-of-Hearing Children. Oxford: Oxford University Press,
pp. 314–343.
Morgan, G. and Woll, B. (eds) (2002), Directions in Sign Language Acquisition Research.
Amsterdam: John Benjamins.
Morgan, G., Barriere, I. and Woll, B. (2003), First verbs in British Sign Language
development. Working Papers in Language and Communication Science, vol. 2,
pp. 57–66.
Newport, E. L., and Meier, R. P. (1985), ‘The acquisition of American Sign Language’,
in D. I. Slobin (ed.), The Crosslinguistic Study of Language Acquisition. Volume 1: The
Data. Hillsdale, NJ: Lawrence Erlbaum Associates, pp. 881–938.
Paul, P. V. and Quigley, S. P. (2000), Language and Deafness. 3rd edition. San Diego, CA:
Singular Publishing Group Inc.
Petitto, L. A. (1992), ‘Modularity and constraints in early lexical acquisition: Evidence
from children’s first words/signs and gestures, in M. R. Gunnar and M. Maratsos
(eds), Modularity and Constraints in Language and Cognition: The Minnesota Symposia
on Child Psychology, Vol. 25. Hillsdale, N J: Lawrence Erlbaum Associates, pp. 25–58.
Pizzuto, E. and Volterra, V. (2000), ‘Iconicity and transparency in sign languages:
A cross-linguistic cross-cultural view’, in K. Emmorey and H. Lane (eds), The Signs of
Language Revisited: An Anthology to Honor Ursula Bellugi and Edward Klima. Mahwah,
NJ: Lawrence Erlbaum Associates, 261–286.
Prinz, P. M., and Prinz, E. A. (1979), ‘Simultaneous acquisition of ASL and spoken
English (in a hearing child of a deaf mother and hearing father). Phase I: Early
lexical development’, Sign Language Studies, 25, 283–296.
Rosenstock, R. (2008), ‘The role of iconicity in international sign’, Sign Language
Studies, 8, 131–159.
Schick, B. S. (1990), ‘The effects of morphosyntactic structure on the acquisition of
classifier predicates in ASL’, in C. Lucas (ed.), Sign Language Research: Theoretical
Issues. Washington, DC: Gallaudet University Press, pp. 358–374.
Schick, B. (2003), ‘The development of ASL and manually-coded English systems’, in

M. Marschark and P. E. Spencer (eds), Oxford Handbook of Deaf Studies, Language,
and Education. New York: Oxford University Press, pp. 219–231.
Singleton, J. L. and Morgan, D. D. (2005), ‘Natural signed language acquisition within
the social context of the classroom’, in B. Schick, M. Marschark and P. Spencer
(eds), Advances in the Sign Language Development of Deaf and Hard-of-Hearing Children.
New York: Oxford University Press, pp. 344–373.
Spencer, P. E. (1993), ‘The expressive communication of hearing mothers and deaf
infants’, American Annals of the Deaf, 138, 275–283.
Stokoe, W. C. (1960), ‘Sign language structure: An outline of the visual communication
system of the American deaf’, Studies in Linguistics: Occasional Paper 8. University of
Buffalo.
Stout, G. F. (1863), A Manual of Psychology. London: University Correspondence
College Press.
Supalla, T. and Webb, R. (1995), ‘The grammar of international sign: A new look at
pidgin languages, in K. Emmorey and J. Reilly (eds), Language, Gesture and Space.
Hillsdale, NJ: Lawrence Erlbaum Associates, pp. 333–352.
Sutton-Spence, R. and Woll, B. (1999), The Linguistics of British Sign Language: An
Introduction. Cambridge: Cambridge University Press.
Tylor, E. B. (1895), Anthropology: An Introduction to the Study of Man and Civilisation.
London: Macmillan.
Van den Bogaerde, B. and Baker, A. (1996), ‘Verbs in the language production of one
deaf and one hearing child of deaf parents. Poster presented at the 5th International
Conference on Theoretical Issues on Sign Language Research, September 1996, Montréal,
Canada.
Volterra, V. and Caselli, C. (1985), From gestures and vocalizations to signs and words.
In W. C. Stokoe and V. Volterra (eds.) The Third International Symposium on Sign
Language Research, Silver Spring: Linstok Press.
Volterra, V., Caselli, M. C., Capirci, O. and Pizzuto, E. (2005), ‘Gesture and the emer-
gence and development of language’, in Tomasello, M., Slobin, D. (eds), Elizabeth
Bates: A Festschrift. Mahwah, NJ: Lawrence Erlbaum Associates, pp. 3–40.
Watson, J. (1809), Instruction of the Deaf and Dumb. London: Darton & Harvey.
Woll, B. (1984), ‘Comparing sign languages’, in F. Loncke, P. Boyes-Braem and Y. Lebrun
(eds), Recent Research on European Sign Languages. Lisse: Swets & Zeitlinger, pp. 79–92.
Woll, B. (1998), ‘Development of signed and spoken languages’, in Gregory, S. and
Powers, S. (eds), Issues in Deaf Education. Clevedon: Multilingual Matters.
Woll, B. (2001), ‘Exploring language, culture and identity: Insights from sign language
and the Deaf community’, in J. Cotterill and A. Ife, (eds), Language across Boundaries.
British Studies in Applied Linguistics 16. London and New York: BAAL (British Associ-
ation for Applied linguistics), in association with Continuum, pp. 65–80.
This page intentionally left blank
Personal Name Index
Archangeli, D. 203, 209 Gunnarsson, B. 7, 123, 124, 125, 127, 128,
129, 130, 132, 134
Ball, M. 8, 202, 207, 209, 213, 215, 217, 219
Bassnett, S. 146, 161 Halliday, M. 165, 166
Bhatia, V.K. 165, 166 Heller, M. 10, 11, 15, 16
Block, D. 2, 11 Hinde, R. 82, 85, 86, 87, 92
Bourdieu, P. 2, 6, 11
Browman, C. 204, 211 Killingley, D. 76, 77
Bush, G.W. 52, 144, 157, 158, 159, 160 Kockelman, P. 83, 84, 85, 86, 88, 89, 92
Bybee, J. 212, 213, 214, 215, 216, 217, Koskinen, K. 145, 147, 150, 154
218, 219
Levinson, S. 83, 86, 88, 89
Cameron, D. 4, 11, 15, 22 Lorch, M. 8, 186, 189, 194, 195, 196, 197
Chomsky, N. 2, 28, 98, 189, 193, 202, 203
Code, C. 209, 216, 217 Meier, R. 225, 230, 232, 235
Coulmas, F. 7, 34, 38 Morgan, G. 8, 230, 232, 234
Coulthard, M. 178, 179 Muller, N. 8, 207, 209, 210
Crystal, D. 4, 48, 115, 166
Nettle, D. 36, 47, 48, 60
de Saussure, F. 28, 100
Dinnsen, D. 204, 210 O’Barr, W. 169, 187
Dodd, B. 209, 218
Dunbar, R. 85, 86, 87, 90, 91, 92 Paradis, M. 188, 189, 190, 191, 192,
193, 195
Eades, D. 169, 171, 176, 179 Pavlenko, A. 7, 10
Enfield, N. 7, 83, 85, 86, 88, 89, 90, 91 Piller, I. 7, 10, 16, 19, 20
Pitres, A. 189, 190, 195, 196
Ferguson, C. 70, 76
Fillmore, C. 101, 165 Robinson, C. 55, 56, 58
Fishman, J. 36, 66, 68 Robinson, C.D.W. 32, 33
Foucault, M. 2, 19 Romaine, S. 7, 36, 47, 48, 56, 58, 59, 60, 61
Rubinstein, A. 29, 30
Gibbons, J. 8, 164, 168, 170, 171
Gierut, J. 208, 210, 215 Sacks, H. 83, 86, 91
Goffman, E. 85, 89, 91, 92, 109 Sawyer, J. 70, 75
Goldstein, L. 204, 211 Schaffner, C. 8, 145, 146, 161
Goodwin, C. 85, 88 Shuy, R. 172, 177
Greenberg, J. 36, 234 Spolsky, B. 7, 65, 66, 68, 71, 72, 73
Grice, P. 31, 89
Grin, F. 11, 16, 32, 35, 36 Tomasello, M. 84, 85, 87
Grunwell, P. 202, 207
Gu, Y. 7, 107, 109, 110, 114, 115, 118, 119 Woll, B. 8, 226, 230, 231, 232, 234, 235, 236
Subject Index
activity zone 105, 107–8, 114 136, 142, 149, 151, 160, 161, 171, 172,
Africa 14, 16, 40, 46, 48, 51, 54, 55, 65, 69, 173, 189, 205, 237
71, 74, 75, 76, 133, 170
age 21, 54, 176, 185, 186, 191, 206, 232, 234 deaf community 89, 168, 223ff
aphasia 185, 187, 189, 190, 191, 195, 229 development 2, 7, 16, 36, 37, 40, 41, 46ff,
Arabic 48, 65, 69, 74–6, 77, 78, 131, 146 70, 71, 72, 74, 76, 77, 85, 86, 88, 108, 135,
Aramaic 66, 67, 69, 75, 76 186, 187, 188, 193, 194, 195, 196, 203,
Asia 14, 16, 74, 75, 76, 77, 133, 164 205, 208, 214, 218, 224, 230, 231, 232,
Australia 20, 22, 49, 50, 54, 61, 73, 167 233, 238
diversity 3, 7, 36, 40, 46ff, 122, 123, 127,
Bible 65, 66, 69, 70, 71, 72 129, 133, 134, 135, 137, 138
bilingual aphasia 189, 190, 191, 192, 194,
195, 229 economic theory 28, 29, 35, 38
bilingualism 2, 4, 8, 12, 13, 14, 22, 78, 130, economy 7, 10, 11, 12, 13, 16, 22, 28ff, 122,
133, 172, 191, 223ff 123, 129, 135, 136, 137, 176
Brazil 50, 52, 128 education 2, 8, 11, 12, 13, 14, 32, 35, 53,
Britain 124, 125, 143, 149, 167, 223 56, 57, 60, 61, 71, 72, 73, 104, 106, 113,
114, 133, 143, 171, 172, 185, 187, 189,
Canada 15, 32, 49, 50, 53, 60, 61, 189 206, 223, 228, 234, 237, 238
child-rearing 12, 21, 22 Egyptian 40, 69, 73
child speech disorder 202, 209, 210 English 10, 11, 13, 14, 15, 17, 18, 20, 22,
China 37, 48, 50, 51, 52, 72, 99, 103, 104, 34, 37, 38, 40, 41, 47, 48, 57, 58, 69, 71,
107, 108, 111, 114, 164 75, 78, 98 122, 123, 124, 127, 128, 129,
Chinese 15, 33, 34, 35, 38, 48, 52, 98, 99, 130, 131, 132, 133, 141, 144, 145, 146,
102, 103, 110, 114, 115, 146, 167 147, 148, 149, 150, 151, 152, 153, 154,
choice 29, 31, 33, 34, 39, 41, 65, 66, 67, 155, 156, 158, 159, 160, 162, 164, 166,
69, 73, 77, 127, 128, 129, 130, 133, 151, 167, 168, 170, 171, 172, 173, 177, 178,
178, 215 185, 189, 193, 195, 202, 205, 207, 208,
Christianity 66, 69, 70, 71, 73, 74 210, 213, 229, 235
clinical data 202, 204, 205, 209, 210, Europe 2, 3, 15, 16, 18, 19, 28, 34, 35, 37,
212, 214 38, 48, 59, 66, 67, 68, 70, 71, 74, 75, 90,
cognitive process 8, 84, 85, 223 124, 129, 133, 143, 144, 145, 146, 147,
colonialism 37, 46, 47 148, 149, 150, 151, 153, 154, 157, 164,
common language 36, 38, 39, 40, 66 167, 172, 177, 229, 236
complexity 33, 39, 84, 122, 150, 161, 167, external communication 122, 126, 130, 135
168, 170, 205
Congo 49, 50, 51, 71, 72 face-to-face interaction 10, 92, 102, 106,
corporate language 122, 130, 131 107, 109, 118, 166
covert contrast 202, 207, 209, 211, 218 forensic linguistics 1, 8, 176, 179
Critical Discourse Analysis (CDA) 6, 122, formality 161, 165, 178
161, 169 French 2, 3, 11, 13, 19, 20, 32, 38, 55,
cross-linguistic 38, 189, 226 56, 67, 71, 78, 98, 128, 142, 144, 146,
culture 7, 35, 47, 57, 59, 60, 72, 75, 83ff, 147, 148, 149, 150, 153, 154, 155, 156,
99, 123, 124, 125, 126, 127, 129, 132, 134, 167, 231
Subject Index 245
GDP (gross domestic product) 34, 36, 50, Latin America 12, 14, 54, 70, 76, 164
57, 59 law 5, 58, 67, 68, 72, 95, 75, 108, 110, 117,
gender 7, 10ff, 36, 169, 176, 188 147, 164ff, 223, 257
genre 21, 135, 145, 161, 165, 166, Lebanon 75, 78, 134
172, 178 lingua franca 13, 72, 74, 123, 129, 130,
German 13, 15, 22, 38, 67, 38, 77, 78, 124, 131, 133, 137
125, 128, 131, 134, 142, 143, 144, 147, local 11, 33, 38, 55, 56, 65, 67, 70, 71, 72,
148, 149, 150, 151, 152, 153, 154, 155, 73, 75, 76, 78, 86, 87, 89, 90, 92, 105, 106,
156, 157, 158, 159, 160, 166 107, 111, 123, 124, 125, 127, 128, 129,
gestural phonology 204, 211, 214, 215, 218 132, 133, 135, 137, 169
global 7, 11, 16, 19, 40, 41, 46, 47, 48, 49,
51, 52, 57, 58, 79, 122, 123, 127, 128, 129, marginalization 56, 129, 134
135, 136, 137, 146 marketing 8, 22, 123, 125, 126
Greek 66, 67, 68, 69, 70, 146, 171 MDG (Millennium Development Goal) 19,
56, 61
Hebrew 18, 65, 66, 67, 68, 72, 77 media 2, 16, 38, 47, 105, 115, 116, 117,
Hindi 48, 71, 75, 77, 146 118, 142, 143, 144, 145, 156, 163,
human rights 58, 59, 61 173, 235
Mexico 12, 50, 52, 54, 78
identity 10, 19, 34, 35, 36, 46, 86, 110, 112, migration 10, 11, 14, 17, 19, 56, 67,
113, 124, 134, 147, 191 114, 186
idiolect 104, 105, 106, 114, 115 minority 12, 13, 14, 15, 16, 18, 21, 22,
immigrant 14, 17, 18, 19, 21, 33, 66, 67, 73, 35, 36, 40, 55, 56, 61, 67, 105, 128, 171,
128, 129, 131, 133, 134, 143, 191 172, 190, 230, 234, 236, 237
India 50, 51, 53, 54, 71, 72, 74, 75, 76, missionary 65, 70, 71, 72, 76
77, 170 mobility 108, 114, 118, 122, 135, 136
indigenous 7, 12, 13, 46, 47, 53, 54, 56, 57, modernization 4, 58, 76, 169
58, 59, 60, 61, 71, 72, 176 Mongolia 52, 104, 107, 108
inequality 16, 38, 47, 51, 52 monolingualism 2, 14, 15, 47, 143
interpersonal 91, 104, 114, 165, 168 motor planning 204, 207, 213, 217
interpreter 15, 39, 76, 89, 157, 170, 171, multilingualism 1, 2, 4, 10ff, 34, 36, 37,
235, 236, 237 122, 129, 130, 189, 192
interpreting 143, 146, 156, 157, 170, multimodality 105, 109, 110, 111, 112,
171, 236 113, 135
intervention 56, 74, 155, 173, 175, 202,
205, 206, 237 native 18, 49, 53, 54, 58, 60, 103, 111, 131,
Islam 65, 66, 74, 75, 76, 79, 162, 166 133, 178, 190, 195, 225, 232, 234
Israel 67, 68, 79, 143, 144, 162 network 5, 10, 20, 35, 39, 41, 46, 83, 112,
113, 122, 136, 137, 147, 191, 212, 213,
Japanese 20, 21, 34, 38, 48, 59, 128, 214, 215, 229
164, 167 New Zealand 61, 70, 71, 73, 138
Judaism 65, 66, 68, 69, 72, 73, 74 Nigeria 13, 49, 50, 51, 53
norm 88, 89, 90, 91, 146, 162
language acquisition 1, 2, 4, 5, 33, 195,
223, 229, 230, 231, 232, 238 Optimality Theory 204, 209, 210
language choice 33, 34, 35, 39, 65, 77, organization 7, 8, 20, 25, 31, 38, 39, 41, 66,
128, 130 72, 74, 84, 85, 122ff, 145, 146, 151, 185,
language shift 13, 37, 60, 172 187, 188, 189, 192, 195, 197, 203, 208,
Latin 65, 69, 70, 73, 167, 168, 175 211, 217, 223, 228
246 Subject Index
Papua New Guinea 46, 50, 53 Spanish 12, 13, 14, 48, 58, 70, 78, 131, 142,
perception 57, 59, 84, 204, 205, 212, 146, 147, 148, 154, 171, 177
214, 228 speech disorder 195, 202, 209, 210, 213
phonological 8, 79, 193, 202ff, 223, 225 speech production 8, 202, 203, 104, 105,
Policy 6, 7, 33, 34, 39, 40, 41, 46, 56, 58, 60, 120, 212, 218
61, 65, 66, 68, 69, 70, 71, 72, 73, 74, 76, spoken discourse 125, 132, 178
77, 110, 127, 129, 134, 137, 145, 146, 147, Sudan 50, 51, 75, 79
148, 152, 175, 234, 238
political 8, 36, 37, 40, 58, 61, 73, 85, 98, translation 3, 8, 38, 39, 65, 66, 67, 68, 69,
107, 118, 142ff 70, 71, 72, 74, 75, 76, 78, 142ff, 170, 171
Portuguese 48, 70, 128, 156, 157 trust 31, 38, 41, 86
poverty 7, 30, 36, 46ff
prayer 65, 66, 67, 69, 78 United Nations (UN) 58, 61, 146
problem-solving 1, 2, 90, 92 United States (US) 14, 19, 32, 46, 49, 50,
psycholinguistics 116, 191, 203, 204, 205, 51, 52, 54, 59, 61, 67, 68, 73, 78, 103, 157,
210, 212, 213, 217, 218 158, 167, 170, 172, 173, 174, 176, 184,
192, 205, 226
Qur’an 74, 75, 76, 78 utility 32, 33, 34, 35, 36, 37, 39, 40
rhetoric 16, 29, 30, 31, 61 vernacular 65, 66, 67, 68, 69, 70, 71, 72, 73,
Russian 18, 48, 52, 70, 73, 78, 131, 75, 77, 78
144, 146
website 106, 127, 128, 130, 146, 151, 152,
school 5, 15, 20, 21, 37, 40, 50, 53, 55, 56, 154, 155, 156, 157, 158, 159, 160, 233
57, 58, 60, 68, 70, 71, 72, 75, 76, 79, 104, well-being 54, 56, 57, 60
114, 133, 173, 185, 187, 191, 203, 224, workplace 111, 114, 118, 122, 123, 124,
226, 228, 234, 238 129, 132, 134, 135, 136, 137, 138
situatedness 105, 107, 110, 111, 112, 118 writing 29, 40, 60, 75, 85, 99, 110, 117, 123,
social relations 6, 7, 83, 86, 91 124, 125, 127, 131, 135, 178, 186, 196, 206
sociolinguistics 66, 115, 122, 134, 191
South-East Asia 18, 75, 164 Yiddish 67, 68, 69, 79

PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PDF

Uploaded by

Copyright:

Available Formats

CONTEMPORARY APPLIED LINGUISTICS

VOLUME 2 LINGUISTICS FOR THE REAL WORLD

Contemporary Corpus Linguistics

Contemporary Applied Linguistics Vols 1 and 2

Volume 2 Linguistics for the Real World

© Li Wei, Vivian Cook and Contributors 2009

British Library Cataloguing-in-Publication Data

ISBN: 978-0-8264-9681-2 (Hardback)

Library of Congress Cataloging-in-Publication Data

Typeset by Newgen Imaging Systems Pvt Ltd, Chennai, India

Notes on Contributors vii

Introduction: Linguistics for the Real World 1

Chapter 11 Clinical Phonology 202

Personal Name Index 243

Martin Ball, Ph.D., mjb0372@louisiana.edu, is Hawthorne-BoRSF Endowed Pro-

Vivian Cook, Ph.D., vivian.cook@ncl.ac.uk, is Professor of Applied Linguistics at

Florian Coulmas, Ph.D., coulmas@dijtokyo.org, is Director of the German Institute

Nick Enfield, Ph.D., nick.enfield@mpi.nl, is a senior staff scientist in the Language

John Gibbons, Ph.D., johgibbo@hotmail.com, was Professor of Linguistics at

Gu Yueguo, Ph.D., yueguo.gu@gmail.com, is Research Professor of Linguistics at

Britt-Louise Gunnarsson, Ph.D., britt-louise.gunnarsson@nordiska.uu.se, is Profes-

Li Wei, Ph.D., li.wei@bbk.ac.uk, is Professor of Applied Linguistics at Birkbeck

Marjorie Lorch, Ph.D., m.lorch@bbk.ac.uk, is Professor of Neurolinguistics at

Gary Morgan, Ph.D., g.morgan@city.ac.uk, is Professor of Psychology at City Uni-

Nicole Müller, Ph.D., nmueller@louisiana.edu, is Professor of Communicative

Aneta Pavlenko, Ph.D., apavlenk@temple.edu, is Professor of TESOL at Temple

Ingrid Piller, Ph.D., ingrid.piller@gmail.com, is the founding director of the

Suzanne Romaine, D.Phil., suzanne.romaine@ling-phil.ox.ac.uk has been Merton

Christina Schäffner, Ph.D., c.schaeffner@aston.ac.uk, is Professor of Translation

Bernard Spolsky, Ph.D., bspolsky@gmail.com, is professor emeritus in the Depart-

Bencie Woll, Ph.D., b.woll@ucl.ac.uk, is Professor of Sign Language and Deaf

applied linguistics is an interdisciplinary field of research and practice dealing

2 Linguistics for the Real World

3 The Present Volume: Individuals Looking to the Future

how an ‘organizational self’ is constructed by means of discourse, and the role of

1.2 Globalization, Multilingualism, and Gender

In conceptualizing ‘multilingualism’, we follow Heller’s (2007a) view of ‘language

begun examining other topics linked to globalization, such as the commodification

changes in recent years, as part of developments that can broadly be termed as

1.3 Gender and Multilingualism in Employment

To better understand the relationship between gender, multilingualism and

1.3.1 Gendered Bilingualism in the ‘Old’ Economic Order

be a Latino immigrant in the United States. Consequently, learning English takes

1.3.2 Gendered Language Work in the “New” Economic Order

a commodity: the commodity being, ironically, a marker of authenticity in tourism

1.3.3 Paid Reproductive Work

In some cases, the language proficiency of reproductive workers may be one of

1.4 Romancing Multilingualism

In addition to work, the family created on the basis of heterosexual coupling is

1.4.1 Language Desire

1.4.2 Multilingual Child-Rearing

Multilingualism is a form of practice, and it is a gendered practice. We have struc-

Smagulova, J. (2008), Kazakhstan: Transformation of Language Ideology. Paper presented

2.1 The Economic Turn

2.2 Economics and Language

2.3 The Language of Economics

2.3.1 The Language of the Profession

Like astronomers, biologists and palaeontologists, economists are concerned with

2.3.2 Optimizing Language

Interest in the professional language of economics is coupled with a more general

2.3.3 Interpretation and Rhetoric

executing transactions, as a means of production, and as a product. In short, lan-

2.4 Economics of Language