Вы находитесь на странице: 1из 258


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 importance to the eld, but also with regards to its wider signi cance either in terms of methodology, practical application or conclusions. The result is a stimulating contemporary snapshot of the eld 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

Notes on Contributors Acknowledgements Introduction: Linguistics for the Real World Li Wei and Vivian Cook Chapter 1 Globalization, Multilingualism, and Gender: Looking into the Future Ingrid Piller and Aneta Pavlenko Language and Economy Florian Coulmas Linguistic Diversity and Poverty: Many Languages and Poor People in a Globalizing World Suzanne Romaine Religious Language Management Bernard Spolsky Language and Culture Nick En eld Four-Borne Discourses: Towards Language as a Multi-Dimensional City Gu Yueguo Discourse in Organizations and Workplaces Britt-Louise Gunnarsson Political Discourse and Translation Christina Schffner Language and the Law John Gibbons Neurolinguistics and the Non-monolingual Brain Marjorie Lorch vii xi 1 10

Chapter 2 Chapter 3

28 46

Chapter 4 Chapter 5 Chapter 6

65 83 98

Chapter 7 Chapter 8 Chapter 9 Chapter 10

122 142 164 184


Contents Clinical Phonology Martin J. Ball and Nicole Mller Sign Linguistics, Sign Language Learning and Sign Bilingualism Gary Morgan and Bencie Woll 202 223

Chapter 11 Chapter 12

Personal Name Index Subject Index

243 244

Notes on Contributors
Martin Ball, Ph.D., mjb0372@louisiana.edu, is Hawthorne-BoRSF Endowed Professor 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 Mller 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 DuisburgEssen, on leave of absence. He has taught and done research in various environments, 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 En eld, Ph.D., nick.en eld@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 Interaction (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).


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 publications 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 Professor 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), Handbook 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 intercultural pragmatics. He is Principal Editor of the International Journal of Bilingualism (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, educators and historians. Gary Morgan, Ph.D., g.morgan@city.ac.uk, is Professor of Psychology at City University 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 nishing a book on sign language learning in the polyglot Savant Christopher Smith, N., Tsimpli, I. M., Morgan, G. and Woll, B. Signs of the Savant. Cambridge University Press. Nicole Mller, 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 bilingualism. Her recent books include Emotions and Multilingualism (Cambridge University Press, 2005), Bilingual Minds (Multilingual Matters, 2006), Crosslinguistic In uence in Language and Cognition (with Scott Jarvis, Routledge, 2008), and Multilingualism 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 University in the United Arab Emirates. Her research is in the sociolinguistics of language learning and intercultural communication. Her books include Bilingual Couples Talk (Benjamins, 2002) and she is currently working on a textbook on Intercultural 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 Troms. She is co-author of Vanishing Voices. The Extinction of the Worlds Languages, which won the British Association of Applied Linguistics book of the year prize in 2001. Christina Schffner, 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 Translation 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 Department 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 Management (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 community, 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).

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 support 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

Li Wei and Vivian Cook

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 (Brum t, 1995: 27). Similarly, the members of the American Association of Applied Linguistics (AAL) promote principled approaches to language-related concerns. The International Association of Applied Linguistics (AILA) proclaims: applied linguistics is an interdisciplinary eld of research and practice dealing with practical problems of language and communication that can be identi ed, analysed or solved by applying available theories, methods or results of Linguistics or by developing new theoretical and methodological frameworks in Linguistics to work on these problems. The AILA de nition is both broader in including more areas and narrower in relating applied linguistics to linguistics proper. If you have a problem with language, send for an applied linguist. The broad de nition of applied linguistics as problem-solving was certainly true in its early days. De nitions of applied linguistics now are more like lists of the areas that make it up. The Cambridge AILA 1969 Congress encompassed rst language acquisition, computational linguistics, forensic linguistics, speech therapy, neurolinguistics, second language acquisition research, and a host more. Gradually many areas have declared unilateral independence from applied linguistics; rst language acquisition research soon disappeared from the fold to found its own organizations, conferences and journals, as did much second language acquisition research slightly later. Applied linguistics gatherings these days are far less inclusive, though there is a growth in the Research Networks such as Multilingualism: Acquisition and Use. The AILA Congress in 2008 had 9 papers on rst language 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 Association (EUROSLA), the International Symposia on Bilingualism (ISB), Generative Approaches to Second Language Acquisition (GASLA), or the International Association for Multilingualism, not to conferences named applied linguistics. Professional 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 re ect 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 something that has gone wrong which can be solved. Talking about the problem of multilingualism, say, is ambiguous between de ning it as a research area and claiming 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 problem 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 regarding 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 AILAs fond belief that linguistics is the core, many feel linguistics is only one of the contributing disciplines. Applied linguists have explored psychological 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 eld 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 eld 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 rede nes 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


and to some extent Jackendoff, the theories come from post-modernism, psychology 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 developments in phonetics and phonology. In a well-regarded book representative of the eld 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 Con serie Franaise (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 methodology or applied sociology or whichever discipline it uses as its source. So what problems does applied linguistics solve? If you are worried about your childs 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 simpli ed 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 seldom be resolved by looking at a single aspect of language. Since applied linguistics 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 lter 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

4 Li Wei and Vivian Cook all the other factors and decide on the solution. Rather than saying You should follow this way of language teaching, the applied linguists 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 interpreting the customers 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 youre doing applied linguistics; describe childrens language or vocabulary and its rst language acquisition; make a description of grammar and youre 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 Beardsmores work with bilingualism. But they have had little impact on public debate or decision-making for most language problems, the honourable exceptions 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 cant 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 nd out the words, then a lipsynch actor to read them in alternative voices till she recognized her fathers. This didnt work until she herself uttered her fathers words. In another scene an elderly aphasic patient delivers a monologue, judging by audience reaction the rst time that most of them had encountered this kind of discourse. At a dinner party, lm actors and agents attempted to converse simultaneously in three languages, to comic effect. Lipsynch movingly showed the importance of language to peoples lives and the language problems they encountered.


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 relevant 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 language 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 challenge both the folk notions of language and grammar and the theoretical linguists 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 at, 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 reconceptualize applied linguistics as part of social science (e.g. Brum t, 1997; Rampton,

6 Li Wei and Vivian Cook 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 speci c contexts. Learners apparent dif culties 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 disappointingly little impact on the broad eld that it aspires to be part of. An admirable 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 analyses 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 support for existing models or policies on a limited scale. There are a number of possible ways forward: one is to come up with some alternative theories of society and social structures. This would, however, risk endangering 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 relationships and social structures. Perhaps the most important way forward for applied linguistics is the development 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 microanalysis. 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 strati ed and social phenomena as related to one another in complex ways.


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 uni ed voice. The contributions do not resemble the genre of book currently called handbooks, which mostly have state-of-the art surveys of the eld or histories of past achievements. Rather the contributors here are individuals laying out their ideas of a future for applied linguistics in its ne 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 intersections, 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 worlds languages but are among the worlds 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 exercise 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 En eld revisits the age-old theme of the relationship between language 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? En elds 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. Gus 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 examines discourse in organizations and workplaces by discussing the discourse-related problems facing managers and employees in the globalized economy. She analyses

8 Li Wei and Vivian Cook how an organizational self is constructed by means of discourse, and the role of discourse for internal management and external marketing. Schffner examines the interface between political discourse analysis and translation studies. She seeks to address the (in)visibility of translation in political communication, textual proles of translations (as compared to their source texts), and the role of the sociopolitical contexts and conditions in which translations are produced. Gibbons in Chapter 9 outlines the central issues in the fast developing eld of Forensic Linguistics, including legal language, language legislation and linguistic 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 signi cance 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 vili cation, 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 linguists have carried out over the years. Lorchs chapter synthesises the key issues emerging from neurolinguistic research. Speci c consideration is given to the development of our understanding of how language is organized in the brain by re nements 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 theoretical issues, methodological strategies and forms of argumentation in neurolinguistics through exemplars from the historical literature of the eld. Ball and Mller in their chapter critically examine the relationship between Clinical Phonology, a sub- eld 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 nal 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 processes 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.


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. Brum t, C. J. (1995), Teacher professionalism and research, in G. Cook and B. Seidlhofer (eds), Principle and Practice in Applied Linguistics. Oxford: Oxford University Press. Brum t, C. J. (1997), How applied linguistics is the same as any other science, International Journal of Applied Linguistics. 7, (1), 8694. 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), 325. 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. 97114. Sealey, A. and Carter, B. (2004), Applied Linguistics as Social Science. London: Continuum.


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 elsewhere (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 clari cation 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 Hellers (2007a) view of language as a set of resources which circulate in unequal ways in social networks and discursive spaces, and whose meaning and value are socially constructed within the constraints of social organizational processes, under speci c historical conditions (2). This view enables us to ask questions about how certain resources come to be valued 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 multilingualism also draws on Appadurais (1990) description of globalization as consisting of ows of goods, capital, communication and people. Many of these ows throw people of widely different linguistic and cultural backgrounds into contact, either virtual as in the case of ows of information, or face-to-face, as in the case of ows of migration and tourism. Applied linguists have widely debated the role of English in the globalization process (Brut-Grif er, 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 commodi cation of language in the new economy (Heller, 2006, 2007b), reordering of local repertoires due to the in ux 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 ows and language education (Block and Cameron, 2002; Kumaravadivelu, 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 (Duchne 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 ows 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 accidental 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 multilingualism 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 Franois 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 pro cient in English reported an average monthly income of CHF 7,636. By contrast, the average monthly income of women who are highly pro cient in English was CHF 4,096. Apart from the shocking difference in net earnings, there was another gender difference: mens earnings rose progressively from no pro ciency in English, via low and basic to uent, i.e. the most pro cient 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 incremental gain associated with uent pro ciency. According to Grin (2001), this non-linearity is due to the fact that pro cient 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 pro cient 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 production 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 difference 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 womens work. We will discuss the ways in which multilingualism can end up as unpaid womens work later in the chapter. At present, in most societies around the world, paid work is no longer the exclusive 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 nd 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 economy and (3) paid reproductive work. We will discuss the ways in which multilingualism 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 agriculture and manufacturing work, became the rst locus of attention of pioneering scholars in the new eld 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 pro cient 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 dif cult for them to join the paid labour force, for which the use of Spanish was crucial, and they

Globalization, Multilingualism and Gender 13 had limited opportunities to practice whatever Spanish they knew. In such cases, a vicious circle may emerge in which limited pro ciency in the majority language limits access to (non-domestic) paid employment, which, in turn, limits interactional 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 language shift to the majority language if they have the opportunity. This pattern was rst 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 signi cant improvement over the drudgery of Hungarian-speaking peasant life. McDonald (1994) reports similar ndings 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 cowshit to nery (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 nd themselves in a doubly marginalized position they are disadvantaged as minority members, and as women. In such circumstances, minority women may misrecognize in Bourdieus (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 subsistence farming. There was also a remarkable difference between the living conditions 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 Nio-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, racialized Quechua speakers. The same awareness does not extend to what it means to

14 Ingrid Piller and Aneta Pavlenko be a Latino immigrant in the United States. Consequently, learning English takes on a greater signi cance 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 cleaning (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 6080 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 particularly 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 employment 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 economic globalization, particularly as far as communities in Africa and Asia are concerned (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 advancement 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 ows of the globalized world enabled some rural minority communities and particularly women in these communities to turn their languages into

Globalization, Multilingualism and Gender 15 a commodity: the commodity being, ironically, a marker of authenticity in tourism contexts. A good example of this process is provided by Nushu (or N Shu), the womens script used by women in Chinas Hunan province. Often used as a textbook example of the only womens language in the world, this script was traditionally used as a secret language by sworn sisters to lament the hardships of having to leave ones 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 signi cant 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 commodied 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 monolingual 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 celebration: 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 workers 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 ts 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 multilingualism, 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 nd economic value in putting their minority language on sale (Watts, 2005). New job opportunities calling for multilingual skills have emerged in the service industry these positions are disproportionately

16 Ingrid Piller and Aneta Pavlenko occupied by women and are also casual, part-time, junior, relatively poorly paid and with little job security. So far, the study of commodi cation of minority languages as signi ers of authenticity in cultural and heritage tourist destinations as well as substantial participation of women in these projects have received only limited sociolinguistic attention (but see Heller (2005) for a notable exception). Future studies need to engage critically with the commodi cation of multilingual skills. There is a need for more industry-speci c ethnographic work considering language work in the service and knowledge economy. Additionally, more quantitative macro-economic investigations in speci c 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 signi cantly more exploitative, and that is paid reproductive work, such as child care, elder care and domestic work. Ehrenreich and Hochschilds (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 adjustment 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 consumerism 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 rst time in recorded history (Ehrenreich and Hochschild, 2002a). Female migratory ows 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 womens 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 af uent 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 children 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

Globalization, Multilingualism and Gender 17 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 pro ciency 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 employers perspective, non-existent or limited pro ciency in the majority language may be an advantage: reproductive work involves access to the most intimate details of family life, and limited pro ciency may serve as a way to keep up the pretence of distance. Limited pro ciency in the majority language may also be one way for the employer to regard the reproductive worker as inferior and to rationalize 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 pro ciency 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 immigration authorities (AAP, 2006). The quote exempli es how crippling limited pro 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 pro ciency 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 emotion 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 dif culty of the work. In the literature we consulted, lack of linguistic pro ciency 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:

18 Ingrid Piller and Aneta Pavlenko I started right away working as a live-in caretaker of an old man. The rst 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 pro ciency 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 bilingual 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 childs 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 russi ed, while the rural populations continued to be monolingual 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 linguistic 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 pro ciencies are valued or not, migrant reproductive workers tend to nd themselves in the same depressing and isolated situation that full-time unpaid homemakers nd themselves 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 linguistic pro ciency 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-byside 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.

Globalization, Multilingualism and Gender 19 The eld also needs longitudinal and socio-economically oriented studies to examine whether and how the mastery of the majority language is converted into economic 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 ows in search of employment are only one kind of transnational ows. Another ow 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 relationship with a speaker of the desired language (possibly to get the coveted legitimacy). The term originates in Pillers (2002) and Piller and Takahashis (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, personalities, life trajectories, or, for that matter, linguistic repertoires. As a result, language desire, too, may be hegemonic and an instrument through which individuals conspire in their own oppression. The rst 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 frivolous Grand Tour tradition (Gore, 2005: 24). The gendered trend continues to this day: according to Kinginger (2008), in 20052006, 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 commodied 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). Kingingers (2008) ethnographic study of American students in France reveals that these dominant ideologies of gender and embodiment are internalized by

20 Ingrid Piller and Aneta Pavlenko many of the study participants yet with different consequences. Men can position themselves as appreciative consumers of stylish French women and bene t from their interactions gaining in linguistic pro ciency (for similar ndings 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 network 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 making almost no contact with French speakers and made no visible linguistic gains by the end of her semester abroad. Importantly, the participants in Kingingers study are middle- and upper-middleclass Westerners on a study-abroad program in another Western country. When racialized non-Westerners come into play, the relations of power change dramatically, as seen in an ethnographic study of Japanese women on a study-abroad experience 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 uent in English easily and effortlessly. When this did not happen and they found that it was actually dif cult to even enter into interactions with English speakers, their search for an English-speaking romantic 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 rst and foremost White, so that the link between symbolic ownership of English and race was reproduced. All of the participants stayed on in Sydney beyond the originally intended period, which led most of them into a nancially dif cult or dependent situation. As the expected rewards of these sacri ces uency in English, full participation in Sydneys 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. Relationship English English for the speci c purpose of conducting a romantic relationship 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 language desire is constituted in public and private discourses in a wider range of contexts, asking whose interests are served or disserved in particular cases.

Globalization, Multilingualism and Gender 21

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 parentchild 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. BarronHauwaert, 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 language 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 genderneutral 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 language work in bilingual parenting. A negative answer to this question comes from Okitas (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 pro ciency 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 childrens preschool years to ensure suf cient 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, intensi ed pressure, guilt and personal trauma (Okita, 2002: 230). Okitas (2002) conclusions can also be extended to parenting in traditional minority and immigrant communities where minority language maintenance is

22 Ingrid Piller and Aneta Pavlenko womens responsibility and, for Cameron (1992), yet another way to keep women within the con nes 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 Schpbach (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 traditional family arrangements.

1.5 Conclusion
Multilingualism is a form of practice, and it is a gendered practice. We have structured 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 multilingualism 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 (commodi ed) 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 languages 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 childrearing as an addition to the numerous invisible tasks of womens work multilingualism 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 bilingualism as if these were intrinsically valuable without reference to their social context is at best nave, 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 communities and individuals.

1 A different version of this paper has appeared as Piller and Pavlenko (2007).

Globalization, Multilingualism and Gender 23

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: Europes domestic workers, Race and Class, 39, (1), 3749. 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, 295310. 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. 1535. Blair-Loy, M. and Jacobs, J. (2003), Globalization, work hours, and the care de cit among stockbrokers. Gender and Society, 17, (2), 230249. 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: Narratives of Displacement. Manchester, UK: St. Jerome Publishing, pp. 128143. Bourdieu, P. (1991), Language and Symbolic Power. Cambridge: Polity. Brut-Grif er, J. (2002), World English: A Study of its Development. Clevedon, UK: Multilingual 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 Commodi cation of Language in the Globalized Service Economy. Journal of Sociolinguistics 4, 3, 323347. Cameron, D. (2005), Communication and commodi cation: Global economic change in sociolinguistic perspective, in Erreygers, G. and G. Jacobs (eds), Language, Communication and the Economy. Amsterdam: John Benjamins, pp. 923. 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), 3259.

24 Ingrid Piller and Aneta Pavlenko

Duchne, 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. 243265. 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. 113. England, K. and Stiell, B. (1997), They think youre as stupid as your English Is: Constructing foreign domestic workers in Toronto. (Citizenship and International Migration). Environment and Planning, A29, (2), 195215. Foucault, M. (1977), Discipline and Punish: The Birth of the Prison. New York: Pantheon. Foucault, M. (1980), Power/Knowledge: Selected Intereviews and Other Writings, 19721977. New York: Pantheon. Friedan, B. (1963), The Feminine Mystique. New York: W. W. Norton. Gal, S. (1978), Peasant men dont get wives: Language and sex roles in a bilingual community, Language in Society, 7, (1), 117. 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 Dont Get Fat: The Secret of Eating for Pleasure. New York: Alfred Knopf. Glucksmann, M. A. (2004), Call con gurations: Varieties of call centre and divisions of labour. Work, Employment and Society, 18, (4), 795811. Gohrisch, J. (1998), Zur Situation des Anglistischen Nachwuchses [the Situation of Junior Academics in English Studies], Anglistik: Mitteilungen des Verbandes Deutscher Anglisten, 9, (2), 131140. Goldstein, T. (2001), Researching womens language practices in multilingual workplaces, in Pavlenko, A., Blackledge, A., Piller, I. & M. Teutsch-Dwyer (eds), Multilingualism, Second Language Learning, and Gender. Berlin/New York: Mouton De Gruyter, pp. 77101. Gordon, D. (2004), Im tired. You clean and cook. Shifting gender identities in second language socialization, TESOL Quarterly, 38, (3), 437457. 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), 6578. 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. 238256. 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 commodi cation of language and identity, Journal of Sociolinguistics, 7, (4), 473492.

Globalization, Multilingualism and Gender 25

Heller, M. (2005), Language, skill and authenticity in the globalized new economy, Noves SL: Revista de Sociolingstica, 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.), Bilingualism: A Social Approach. Hampshire/New York: Palgrave, pp. 122. 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. 287304. Hill, J. (1987), Womens 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. 121160. Hochschild, A. (2003), The Second Shift. New York: Penguin. Hondagneu-Sotelo, P. (2001), Domstica: Immigrant Workers Cleaning and Caring in the Shadows of Af uence. 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. 307310. 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. 85110. Moyer, M. and L. Martin-Rojo (2007), Language, migration, and citizenship: New challenges in the regulation of bilingualism, in Heller, M. (ed.), Bilingualism: A social Approach. Hampshire/New York: Palgrave, pp. 137160. Munck, R. (2005), Social exclusion: New inequality paradigm for the era of globalization? In Romero, M. and E. Margolis (eds) The Blackwell Companion to Social Inequalities. Oxford: Blackwell, pp. 3149. Nio-Murcia, M. (2003), English is like the dollar: Hard currency ideology and the status of English in Peru, World Englishes, 22, (2), 121141. 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 Intermarried Families. Amsterdam: Benjamins. Ollivier, D. (2003), Entre Nous: A Womans Guide to Finding Her Inner French Girl. New York: St. Martins Grif n. 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.

26 Ingrid Piller and Aneta Pavlenko

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. 1752. 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. 5769. Pennycook, A. (1994), The Cultural Politics of English as an International Language. London: Longman. Pennycook, A. (2007), Global Englishes and transcultural ows. 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. 341359. Piller, I. and Pavlenko, A. (2004), Bilingualism and gender, in Bhatia, T. and W. Ritchie (eds), The Handbook of Bilingualism. Oxford: Blackwell, pp. 489511. Piller, I. and Pavlenko, A. (2007), Globalization, Gender, and Multilingualism, in Decke-Cornhill, H. and L. Volkmann (eds), Gender Studies and Foreign Language Teaching. Tu bingen, Germany: Gunter Narr Verlag, pp. 1530. 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. 5983. Polanyi, L. (1995), Language learning and living abroad: Stories from the eld, in Freed, B. (ed.), Second Language Acquisition in a Study Abroad Context. Amsterdam/ Philadelphia, PA: John Benjamins Publishing Company, pp. 271291. 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. 7195. 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), 727749. 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. 269287. Schlieben-Lange, B. (1977), The language situation in Southern France. Linguistics, 19 101108. Schpbach, D. (2005), Shared languages, shared identities, shared stories? A qualitative 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.

Globalization, Multilingualism and Gender 27

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 Globalization. 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. 3043. 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), 261277. 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.


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 ve 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 pro t maximization, competition and capital accumulation have fanned out from the narrow con nes 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 behaviour, 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 economists. 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, sociologists 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 understand 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 speci c interest in and approach to the study of language. This variety re ects the centrality of language to human existence. Unlike the elds just mentioned, econo-linguistics is not a recognized sub-discipline of the language sciences, nor is there a uni ed body of knowledge that warrants designation 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 interest 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 economic 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 elds 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 ef ciency, 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 discussion about the methodology of economics has revolved around the rhetoric and style of economic writing. Behind the ensuing interest by economists in analytic 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 knowledge 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 information available to the agent. Treating information as objective, rather than being an excusable simpli cation, is a serious aw 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 interpreted 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 identi ed the power of language as a major force determining the success or failure of economic theories. To these economists, linguistic determinism, especially the metaphors-welive-by strand, has a strong appeal (Gramm, 1996; Levy, 1997). The metaphors of economics, they hold, can in uence behaviour and attitudes, including those of the scienti c 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 reality, 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 theorys 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) rst 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 ful l its communitys 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 de ned. Meanwhile, economists 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 preference), 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 underspeci cation 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 behaviour affects action (Samuels, 1990). On this issue too, there is the economists

Language and Economy 31 perspective and the economic actors perspective. Both the former and the latter habitually spend time and effort to communicate their ideas with the purpose of in uencing 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 rhetoric in this context is the subject of much of McCloskeys (1983) work, upon which others have built (e.g. Klamer, 2001, Mki, 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 imsily 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 humanities than McCloskey, who has said of economists that they are poets without knowing 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, economics must be reconstructed as rhetoric the analysis of persuasive talk which enables inventors to raise capital, entrepreneurs and labour to negotiate and cooperate, advertisers to appeal to customers, stockbrokers to talk to clients and to each other and so on (McCloskey, 1994: 373374). In short, economic actors talk with each other, a fact that has not been lost on economists. As Alvesson and Krreman (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 philosopher of language Paul Grice (1975) as the basis of verbal interaction, has therefore 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 economic texts and in their attempt to assess the role of language in the economy include literary theory, rhetoric and speech act theory. As mentioned, McCloskeys 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, language 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 McCloskeys approach is to focus attention on language not just as a tool that economists use, but as a phenomenon that affects economic processes in multiple 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 de ned either in Marxian terms of control over the means of production, or along Weberian lines relating to the assets which individuals bring to the market and which determine their life chances. Assets considered 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 of cial 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 (Migu, 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 Christo des 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 rst languages and that of subsequently acquired languages does not yield the same rate of return on investment. The same is true

Language and Economy 33 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 speci c 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 ef ciency is a guiding principle 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 dif culty. In a study on the costs and bene ts of language policies in Quebec, Vaillancourt (1985, 1987) therefore proposed a methodology that makes a distinction between quanti able costs and non-quanti able psychic costs. Analogously quanti able and non-quanti able 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 expenditures and returns must be reckoned with. While little is known about the empirical weight of this type of expenditure and return, as compared to quanti able costs and bene ts, 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 assumptions, 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 languages 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 conspicuous 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 Robinsons 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 economic valuation (Vaillancourt, 1991; Coulmas, 1992). Various scales of measurement 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 language as a foreign language (Ammon, 2003). Relating GDP and language illuminates one aspect of the economic signi cance 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 surprisingly, 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. Multilingualism in many parts of the world obscures the picture and in some cases Hong Kong for example makes it dif cult 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 insigni cant languages in a world increasingly subservient to the imperatives of utility, pro t maximization and capital 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-de ned standard of utility 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 in uenced in their choices by instrumental as well as symbolic concerns. Nationalism provides the motivation to choose i.e., legally protect, institutionally support and nancially promote a certain language for national and/or of cial functions (often the default choice). By contrast, the choice of languages for foreign language curricula is more likely to be in uenced 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.

Language and Economy 35 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 bene cial 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 bene ts 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 externalities can be identi ed. 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 languages utility, the more likely it is to attract new speakers. Whether there is a threshold and, if there is, whether it is quanti able, 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 communitys culture and ideologies which may have corrosive and even destructive effects on other communities. This perspective is of particular relevance in minority situations, for while learning a language does not imply unlearning another, the language 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 bene cial to that community, while the bene ts generated 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 of cial 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 languages utility, 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 recognized 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 presuppositions of a truly free and all-encompassing trade commensurate with modern capitalism (Lenin, 1980: 398). A big market promises higher pro ts; 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 in uential studies on this correlation by Fishman (1968) and Pool (1972) found that linguistic fragmentation is detrimental to economic development. 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 expectancy, 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 linguistically relatively homogeneous countries could be poor, heterogeneous ones could never be rich. Nettle (2000) has re-examined these ndings 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 development (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 nd the optimal answer to the question of how people should communicate with each other. By contrast, defenders of minority rights who, together with the instrumental function of language, emphasize its symbolic function for identity and recognize diversity as inherently valuable conceptualize multilingualism as an asset deserving of state protection (Walsh, 2006).

Language and Economy 37 Isolating the factor of language remains dif cult, and an unequivocal correlation 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 capitalist system, the more readily small speech communities have moved away from their languages because parents see better life chances for their children in languages associated with commerce and business opportunity. Maximizing ones communication opportunities to improve employment chances is a strong force behind language shift, a process that seems to be accelerating in the wake of globalization 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 Commodi cation 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 traf c of goods, services and people; an increasingly mobile reserve army of migrant labour; mass tourism; and the communications revolution. Language itself has become the mainstay of a multibillion dollar industry. Selling language is big business. The economic inequality of languages nds vivid expression in the language market. The vast majority of the worlds languages dont sell at all, while a handful of European languages augmented 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 pro ts 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 pro ts they yield, reliable statistics are scarce and estimates on the size of the world market accordingly dif cult. One thing is certain: since McCallens study was published this market has expanded further, especially in Eastern Europe and China (Hildebrandt and Liu, 1991) after Soviet political in uence waned. It is obvious then that an economic valuation of languages 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 notwithstanding, 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 signi 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 contrast 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) identied the heart of this analogy as a human convention without any promise, dwelling on the same crucial property of language mentioned above in connection with the cooperative principle of conversation and its signi cance 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 overview 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 rms activities 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 insuf cient 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

Language and Economy 39 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 interpreters, buy automatic translation software, or outsource language work? More and more internationally active businesses have to nd answers to these questions, answers that will meet their speci c needs. Failure to do so can be costly. Economists 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 ve 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 re ects 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 complicated 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 speci c purposes and thus contributes to a better understanding of the role language plays in the creation and management of organizational capital.

2.4.8 Language as a Public Good

For trading partners it is bene cial to share a common language, as indeed it is bene cial 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 bene ts from its use. And there is no good reason to want to exclude anyone from using it, for one actors use does not reduce the amount that is left for others. These are the de ning 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 languages 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 considered a natural endowment that de nes 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 without 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 rst 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 language or languages. An example is the Pan South African Language Board, which is responsible for the countrys eleven of cial 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 dependent 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, languages 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 bene ts 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 arti cial 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 moribund languages spoken by shrinking communities are characterized by reduced expressive power. Both kinds of languages are found in language contact situations only where competing languages assume some of the functions of the public good. Is it desirable and feasible to in uence 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?

Language and Economy 41 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 linguistic 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 optimalityseeking invisible hand is a worthy project in itself, though one that must be the subject of another chapter.

Alvesson, M. and Krreman, D. (2000), Taking the linguistic turn in organizational research: Challenges, responses, consequences, The Journal of Applied Behavioral Science, 36, (2), 135158. Ammon, U. (2003), The international standing of the German language, in Jacques Maurais and Michael A. Morris (eds), Languages in a Globalising World. Cambridge: Cambridge University Press, pp. 231248. Becker, G. S. (1975), Human capital. A Theoretical and Empirical Analysis with Special Reference to Education. New York: Columbia University Press. (Prcis: http://www. econlib.org/library/Enc/HumanCapital.html), accessed 11.09.2009. Bloom, D. E. and Grenier, G. (1992), Economic perspectives on language: The relative value of bilingualism in Canada and the United States, in James Crawford (ed.), A Source Book on the Of cial Language Controversy. Chicago: University of Chicago Press, pp. 445452. 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, 4568. Breton, A. (1978), Nationalism and language policies, Canadian Journal of Economics, 11, 656668. 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. 133. 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. 261273. Calvet, L.-J. (1984), Troc, march et change linguistique, Langage et Socit, 27, 5581. Chiswick, B. (1978), The effect of Americanization on the earnings of foreign-born men, Journal of Political Economy, 85, 897921. Christo des, 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. 121185. 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 international trade proceedings, Journal of Asian Paci c Communication, 2, (1), 103116. 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. 6387. 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), Lcriture 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, 135. Durocher, R. and Linteau, P.-A. (1971), Le retard de Qubec et linferiorit conomique des Canadiens franais. Montreal: ditions du Boral. Ferraro, F., Pfeffer, J. and Sutton, R. I. (2005), Economics language and assumptions: How theories can become self-ful lling, Academy of Management Review, 30, (1), 824. Fishman, J. A. (1968), Some contrasts between linguistically homogeneous and linguistically heterogeneous polities, in J. A. Fishman, C. A. Ferguson and J. Das Gupta (eds), Language Problems of Developing Nations. New York: Wiley, pp. 5368. Fixman, C. S. (1990), The foreign language needs of U.K.-based corporations, Annals of the American Academy of Political and Social Science, 511, 2546. Gauthier, V., Didiot-Cook, H. and Scheirlinckx, K. (2007), Language needs in business, 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. 147170. Greenberg, J. H. (1956), The measurement of linguistic diversity. Language, 32, 109115. 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. 4158.

Language and Economy 43

Grin, F. (1993), The relevance of thresholds in language maintenance and shift: A theoretical examination, Journal of Multilingual and Multicultural Development, 14, 375392. Grin, F. (1996), Economic approaches to language and language planning: An introduction, International Journal of the Sociology of Language, 121, 116. Grin, F. (1999), Comptences et rcompensenses. 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), 519530. 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 Paci c Communication, 2, (1), 4567. Hocevar, T. (1983), Les aspects conomiques de la dynamique fonctionelle des langues. Language Problems and Language Planning, 7, 135147. Holden, N. (1987), Language barriers as differential constraints on the international behavior of rms, 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. 119137. Hume, D. (1964), Of the origin of justice and property, in A Treatise of Human Nature. Aalen: Scientia, Book 3, 258273 (reprint of second edition, London, 1886). Ingham, G. (2006), The nature of money. American Journal of Sociology, 111, (4), 12271229. Katz, M. L. and Shapiro, C. (1985), Network externalities, competition, and compatibility, American Economic Review, 75, 424440. Klamer, A. (2001), Making sense of economists: From falsi cation to rhetoric and beyond, Journal of Economic Methodology, 8, 6975. 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), 95126. Legrain, P. (2006), Why NAMA liberalisation is good for developing countries, The World Economy, 29, (10), 13491362. Lenin, W. I. (1980), ber das Selbstbestimmungsrecht der Nationen. In: Ausgewhlte Werke. Berlin: Dietz Verlag, 2, 393463. Levy, D. M. (1997), Adam Smiths rational choice linguistics. Economic Inquiry, 35, 672678.

44 Florian Coulmas
Mki, U. (2000), Performance against dialogue, or answering and really answering: A participant observers re ections on the McCloskey conversation, Journal of Economic Issues, 34, 4359. Marschak, J. (1965), Economics of language, Behavioral Science, 10, 135140. Marschan-Piekkari, R., Welch, D. and Welch, L. (1999), In the shadow: The impact of language on structure, power and communication in the multinational, International Business Review, 8, 421440. 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, 481517. McCloskey, D. N. (1990), If Youre 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 materials, Journal of Asian Paci c Communication, 2, (1), 185189. 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, 191195. Metcalf, H. (1992), Foreign Language Needs of Business. Falmer: University of Sussex, Institute of Manpower Studies (Report No. 215). Migu, J.-L. (1970), Le nationalism, lunit nationale et la thorie conomique de linformation, Revue Canadienne dconomique, 3, 183198. Nettle, D. (2000), Linguistic fragmentation and the wealth of nations: The FishmanPool hypothesis reexamined, Economic Development and Cultural Change, 48, 335348. 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. 213230. Qizilbash, M. (2003), Vague language and precise measurement: The case of poverty, Journal of Economic Methodology, 10, (1), 4158. Riedler, N. and Pons-Riedler, S. (1986), An economic analysis of Canadian language policies: A model and its implementation, Language Problems and Language Planning, 10, 4258. Robinson, C. (1988), Language choice: The distribution of language skills and earnings in a dual language economy, Research in Labor Economics, 9, 5390. 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 University Press. Sally, D. (2004), Can I say bobobo and mean Theres no such thing as cheap talk?, Journal of Economic Behavior and Organization, 245266. Samuels, W. J. (1990), Economics as Discourse: An Analysis of the Language of Economists. Boston: Kluwer Academic.

Language and Economy 45

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 franaise, n 20). Quebec: Editeur Of ciel du Qubec. Vaillancourt, F. (1987), The costs and bene ts of language policy in Qubec, 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. 7193. Vaillancourt, F. (1991), The economics of language: Theory, empiricism and application to the Asian Paci c, Journal of Asian Paci c Communication, 2, (1), 2967. Walsh, J. (2006), Language and socio-economic development: Towards a theoretical framework, Language Problems and Language Planning, 30, 127148. Zipf, George K. (1949), Human Behavior and the Principle of Least Effort: An Introduction to Human Ecology. Cambridge, MA: Addison-Wesley Press.


3 Linguistic Diversity and Poverty: Many Languages and Poor People in a Globalizing World
Suzanne Romaine

3.1 Introduction
Many of the worlds poorest people live in regions rich in linguistic diversity. By almost any measure, e.g., the African continent (and sub-Saharan Africa in particular) is high in both linguistic diversity and poverty. Africa as a whole is home to around 2,092 (30.3 per cent) of the worlds 6,912 languages. In 2006 more than a third of the worlds poor lived in Africa, where mean income has been falling since 1981. Sub-Saharan Africa has the highest incidence of extreme poverty and greatest 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 worlds poorest. Its 820 languages spoken by only 0.1 per cent of the worlds population, occupying 0.4 per cent of the worlds land area, comprise ca. 12 per cent of the worlds languages. The overall 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 development economics, the connections between them have not been explored, much less understood. This chapter examines the interface between poverty and linguistic diversity by looking rst at the distribution of global patterns of linguistic diversity, 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 worlds languages but are among the worlds poorest. Section 3.7 explores the policy implications of the overlap between linguistic diversity and poverty in todays globalizing world. My analysis shows how dominant discourses are embedded within a network of systems of knowledge and power that in uence how people think about development, poverty and linguistic diversity. They construct a new world order of international 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 economic 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 bene cial to the world, which has freely chosen the language due to its inherent advantages. Within this worldview, diversity has been problematized while monolingualism or linguistic uniformity is normalized. Just as development and modernization have been taken to be synonymous with the introduction of western science, technology and languages, it is still widely believed that indigenous languages are not suited for modern purposes (Romaine, 2006). Such beliefs have prompted development strategies that result in hegemonic monocultures. Although the fate of the worlds linguistic diversity has not gured 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 worlds 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 abandonment of traditional languages and cultures as an inevitable prerequisite to modernization, 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 communities 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, colonialism, later the Industrial Revolution and today the globalization of economies and mass media (particularly the internet) have created the global village phenomenon. 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 speakers, spoken by just over 41 per cent of the worlds population. These gures include only rst 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
Chinese [Mandarin] Spanish English Arabic Hindi Portuguese Bengali Russian Japanese Total

873,000,000 322,000,000 309,000,000 206,000,000 181,000,000 177,000,000 171,000,000 145,000,000 122,000,000 2,506,000,000

% of World population
14.5 5.4 5.1 3.4 3.0 2.9 2.8 2.4 2.2 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 elsewhere 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 regions complex topography helped to maintain the distinctiveness 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 worlds 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 languages of the world is only 5,0006,000. Nearly 80 per cent of the worlds population speaks a total of only 83 languages, each with 10 million or more speakers. Leaving aside the worlds 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 worlds population. Most, if not all, of these may be at risk, because their speakers are among the worlds most marginalized peoples, who have generally been on the retreat for several hundred years. Speakers of large languages like English and Chinese would nd it dif cult to imagine the prospect of being the last speakers of their language, but the last speakers of probably half the worlds languages are alive today. Crystal (2000: 19) suggests that one language may vanish every two weeks. An increasing number of

Linguistic Diversity and Poverty 49


3,586 smallest languages


spoken by

2,935 mid-size languages

20.4% spoken by


spoken by

79.4% of population

Figure 3.1 Relationship between languages and speakers.

elderly people nd themselves in the position of Jim Andreas, the last known person 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 worlds languages. A further 344 languages 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 uently 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 worlds languages are found in just 20 nationstates, 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 languages in relation to their ranking on the United Nations HDI (Human Development 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 dimensions 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 benets 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 Nations 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 compiled from Gordon 2005 and Human Development Report, 2007/2008: 356.). Country N languages % of worlds languages
11.86 10.73 7.47 6.18 4.5 4.3 4.05 3.98 3.49 3.12 2.89 2.6 2.13 2.1 1.94 1.92 1.87 1.85 1.81 1.66 80.45

N speakers

HDI rank

Papua New Guinea Indonesia Nigeria India United States Mexico Cameroon Australia China Democratic Republic of the Congo Brazil Philippines Malaysia Canada Sudan Chad Russia Tanzania Nepal Vanuatu TOTAL

820 742 516 427 311 297 280 275 241 216 200 180 147 145 134 133 129 128 125 115 5561

3,665,383 218,607,876 90,026,548 943,283,395 262,740,392 92,701,909 9,638,055 16,781,815 1,226,606,396 37,945,510 165,836,223 70,556,507 15,676,135 27,388,756 23,492,482 5,909,406 139,356,854 25,837,668 22,983,905 119,759 3,399,154,974

145 107 158 128 12 52 144 3 81 168 70 90 63 4 147 170 67 159 142 120

Linguistic Diversity and Poverty 51 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 exception, sub-Saharan Africa, which has stagnated since the early 1990s. The bottom 24 countries are all African. The bottom ve 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 ve 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) identi ed as one of the central human development challenges in the decades ahead the need to diminish the tolerance for extreme inequalities that have characterized globalization since the early 1990s and to ensure that the rising tide of prosperity extends opportunities for the many, and not just the privileged few. The poorest one- fth 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 regions 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 worlds top three billionaires amount to more than the combined gross national products of all the least developed countries and their 600 million people. Such disparities indicate that wealth accumulation at the top of the global income distribution has been more impressive 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 speci c, it is socially speci c. Closer examination 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 Canadas 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 (de ned 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 confront 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 de ned 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 prosperous 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 comparable 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
Brazil China Mexico Russia United States

HDI rank
63 85 53 62 10

Poorest 20%
115 115 108 95 31

52 30 55 33 21

Linguistic Diversity and Poverty 53

3.4 The Poorest of the Poor: Poverty and Linguistic Diversity among Indigenous Peoples
The situation of the worlds 370 some million indigenous and tribal peoples is typically not re ected in statistics for a variety of reasons or is hidden by national averages such as HDI. Indigenous peoples are often the most marginalized peoples in society, without adequate access to education, health care and water. They are frequently 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 of cial statistics because they are a very small proportion, less than 3 per cent, of the countrys 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 Australians is 17 years lower than for non-Aboriginals, which puts them on a par with people 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 ranking, the differences between native people (i.e. First Nations, Mtis 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 gures 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 selfinjury 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 identi able 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 countries 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 absolute 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 bene t less from poverty reduction efforts. The same is true for some of Indias 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 worlds twenty most linguistically 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 outside the forest in xed settlements are unable to meet their food needs and experience higher rates of infectious diseases such as malaria, tuberculosis, HIV/AIDS and parasites. In the absence of traditional cultural practices reducing social tensions, domestic violence against women and alcohol abuse have increased (Ohenjo et al., 2006: 19391941). 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.

Linguistic Diversity and Poverty 55 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 nations 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 justi ed its relocation of San people from the Central Kalahari Game Reserve by claiming that the San deplete the natural 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 increasingly 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 neighbours. 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 delivered, is seen as belonging to of cial 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: 212213). The educational system relies exclusively on French. Thus, development encounters exacerbate a self- perpetuating cycle involving language, development and education. Just as children cannot bene t from schooling in a language they do not understand, the rural poor cannot bene t from development assistance rendered in a language they do not understand. In most parts of the world schooling is still virtually synonymous with learning a second language, particularly in Africa, where few African languages are used in government or education. Robinson (1996: 219223) shows how language use is intimately connected with the exercise and distribution of power and control in meetings held by development agents and other of cials. Where French is used, the agenda and discussion remains outside the control of villagers. One of cial 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 relying on French, development activities and resources reach those who are relatively well educated and whose access to resources is already higher, all the while marginalizing the less advantaged still further. Despite the widespread recognition of womens lower levels of literacy, access to schooling, etc., development does not address the fact that speaking a minority language compounds the marginalization 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 asco (Romaine, 1990). Yet few have asked the rural poor what their conceptions and understanding of development 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 communities, 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 societys 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 obstacle 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 participation in the mainstream urban economy as the core means to enhance well-being. Many indigenous people de ne themselves in terms of their close relationships to land and community, where a good life is associated with maintaining traditional hunting, gathering and herding practices. Many of the Arctics residents,

Linguistic Diversity and Poverty 57 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 living in material terms (AHDR, 2004: 1617). 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 indigenous 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 populations in terms of economic measures, other factors such as discrimination and the signi cance of indigenous perceptions and priorities are often ignored and/or discounted. Wealthy nations continue to set the agenda for development, de ne 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 worlds 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 worlds most important language of the post-industrial global village, English is seen as the epitome of a modern language, the road to development, 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 Chiles educational program to ensure that all students graduating from high school are uent in the globes international language. Chiles minister of education was reported as saying we know our lives are linked more and more to an international presence, and if you cant speak English, you cant sell and you cant 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 ner 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 de ne and defend Chilean culture. Poverty and isolation are not cultural protection. Chiles 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 globes 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 speci cally mentioned, except in the editorials 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 Chiles 600,000 Mapuche people, who account for about 4 per cent of the population, 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 Martnez-Fernndez (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 ttest. If size matters, the bigger the better. These ideas t 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 modernization 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 Commissioner for Human Rights, Mary Robinson made explicit the linkage between poverty 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 de ned 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 Declaration of Human Rights, these consist of both civil and political rights as well as economic, 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 paradoxically 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 Programs (2000: 8) characterization of poverty

Linguistic Diversity and Poverty 59 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 disadvantaged areas and populations does not directly address the socio-economic and cultural processes and policies that lead to inequalities and poverty in the rst place. Despite the rights-conscious approach articulated in the Human Development Reports, which stress that human rights are not optional extras, but rather binding obligations that re ect 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 nd 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 Indigenous 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, af rmative action and special measures for indigenous peoples along with safeguards ensuring their participation 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 activities 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 speci c link between indigenous language loss and communitylevel 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 exceptionless 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 maintenance matters too. Most importantly, these studies provide strong empirical evidence against the view that assimilation to dominant cultures in the interests of modernity is harmless or even bene cial 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 endangered languages as well as the smallest language populations, the ndings are clearly relevant not only for other native communities in Canada but for indigenous peoples everywhere experiencing language shift. According to the 2001 Canadian Census, overall only 15 per cent of the countrys Aboriginal children learn an indigenous mother tongue, and fewer still are spoken to in such a language at home. Global estimates for language loss range from 50 to 90 per cent of the worlds 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 communities 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 knowledge 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

Linguistic Diversity and Poverty 61 activities . . . as part of a process which is intimately linked to the quest for selfdetermination 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 overcome indigenous poverty. Self-determination and self-government are profoundly connected, and public policy has to take this into account. Despite this convergence of ndings 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 indigenous 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 development down increasingly separate paths. The rhetoric portraying western models of government, development and education as the only viable ones for contemporary societies continues the colonizing agenda of the past. Rejecting that colonization entails rejecting the discourse as well. The nation-state remains the most critical unit of analysis because it is the policies pursued within national boundaries 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 regulate language policies exist within the political bodies of individual states makes planning and policy at the supranational level dif cult. 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 worlds languages goes hand in hand with achieving and maintaining greater self-determination as part of a larger strategy of cultural survival and a much larger healing journey.

Alter, J. (2005), Poverty in America: The truths that Katrina laid bare. Newsweek, CXLVI, (12), 4249. 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 worlds poorest fared since the early 1980s?, The World Bank Research Observer, 19,(2),141169. 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),47. Deaton, A. (2006), Measuring poverty, in A. Banerjee, R. Benabou, and D. Mookerjee (eds), Understanding Poverty. Oxford: Oxford University Press, pp. 316. 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 knowledge and youth suicide, Cognitive Development, 22,(3), 392399. Health Canada (2005), A Statistical Pro le 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 Worlds Languages. London: Routledge, pp. 34-41. Kirmayer, L. J. (1994), Suicide among Canadian aboriginal people, Transcultural Psychiatric Research Review, 31, 357. Lalonde, C. E. (2006), Identity formation and cultural resilience in Aboriginal communities, in R. J. Flynn, P. Dudding and J. G. Barber (eds), Promoting Resilient Development in Young People Receiving Care: International Perspectives on Theory, Research, Practice and Policy. Ottawa: University of Ottawa Press, pp. 5271. Martnez-Fernndez, 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 document, accessed 21 November 2007. DOI: 10.1371/journal.pmed.0030260. Nettle, D. and Romaine, S. (2000), Vanishing Voices. The Extinction of the Worlds Languages. Oxford: Oxford University Press.

Linguistic Diversity and Poverty 63

Ohenjo, N., Willis, R., Jackson, D., Nettleton, C., Good, K. and Mugarura, B. (2006), Health of indigenous people in Africa, The Lancet, 367, (9526), 19371946. Patel, R. N., Wallace, L. J. D. and Paulozzi, L. (2005), Atlas of Injury Mortality among American Indian and Alaska Native Children and Youth, 19891998. 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), 443475. 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. 217236. 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. 127146. 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, 8596. 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. 721. Tauli-Corpuz, V. (2006), Address to the opening of the fth session of the Permanent Forum on Indigenous Issues. New York. 15 May. Electronic document, accessed 20 November 2007. http://www.un.org/esa/socdev/unp i/en/session_ fth.html Taylor, J. (2007), Indigenous peoples and indicators of well-being: Australian perspectives on United Nations global frameworks, Social Indicators Research, 82, 116. 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 Development Program. United Nations Development Program (2007/2008), Human Development Report. Fighting Climate Change: Human Solidarity in a Divided World. New York: United Nations Development Program. United Nations Permanent Forum on Indigenous Issues (2005), Millennium Development 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/unp i/en/session_ fourth.html United Nations Permanent Forum on Indigenous Issues (2006), Millennium Development Goals and Indigenous peoples: Rede ning the Millennium Development Goals. Electronic document, accessed 20 November 2007. http://www.un.org/esa/socdev/unp i/ en/session_ fth.html Wall Street Journal (2004), Se habla ingles, 30 December. Wilmer, F. (1993), The Indigenous Voice in World Politics. Newbury Park: Sage.


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 decisions 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 eld has not received the attention it deserves. Religious institutions have been the focus of language con ict. One of the issues of dispute between Reform and Orthodox Judaism in the nineteenth and twentieth centuries was the language of prayer. The Vatican decision to use the vernacular 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 institutions 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 rm 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 Presidents prayer breakfast, to realize this. For adherents, a religious institution is commonly the rst social structure outside the family that aims to in uence 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 immigrants, 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 leaders impinge on language practices and beliefs. I rst discuss the language policy, beliefs and management efforts associated with speci c religions, and then derive some general principles. There are two major questions: what success might religious 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 Fishmans (1972) warning about the need to de ne 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 language 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 Enlightenment 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 tendency 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 rst that in the major rst 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 uniformity 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 translation 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 pro ciency 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 ve, 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 communities continued to use a traditional Ashkenazi pronunciation, but slowly many accepted the modi ed 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 communities, 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 neighbourhoods, 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 in uence 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

Religious Language Management 69 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, learning 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 incorruptible 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 language 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 languages. 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 original 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 of cial religion of the Roman Empire by the end of the fourth century. The Vulgate translation into Latin of the Bible in the fth 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 Spanish and Portuguese conquests of South America were partly conceived and operated 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 missionaries started to spread ahead of or together with the soldiers and sailors establishing 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 satis ed 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 literacy 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 missionary work among the Polynesians in the South Paci c. In Samoa, New Zealand and Tonga, English missionaries arriving at the beginning of the nineteenth century 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 children and adults to read and write in the Tongan language (Latukefu, 1974, 1980: 55).

Religious Language Management 71 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 rst 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 proli c of the translators was J. E. Moulton, who translated two volumes of world history, Miltons Paradise Lost, two volumes of Pilgrims 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 Maori vernacular literacy, so that by 1860, there was probably higher literacy among the indigenous Maori 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 Maori 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 Maori. 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 IIs 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 of cial documents but allowed the option of teaching African languages 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 masspublication, 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 assimilationists 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 rst, 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 in uence on the development of limited vernacular literacy. The in uence 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 owering 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 written 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 without note or comment. In its rst 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 rst translation (in Delaware) two years later. It supported work in India and China and the Levant. Founded in 1934 as a summer 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 eldwork studying the language 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 de nes itself as a non-pro t, scienti c 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

Religious Language Management 73 of ritual and adding an educational system to guarantee intergenerational continuity, 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 signi cant role in supporting family language policy in the encouragement of immigrant language maintenance. A signi 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 in uence (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 honori cs 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 of cial 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 speci c ruling. In Christianity, it depends on the organization of the branch of the church, as to whether an individual or de ned 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 premodern period.13 In language policy, Islam broke from both its models in its continuing insistence on the higher status of the Classical Arabic in which the Quran was composed. The Quran, 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 Quran contains the precise words of God, without human intervention 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 religion and of Arabic, noted not just in ordinary speech but also in recitation of the Quran, where it represented a dangerous interference in the effort to ascertain of the message of the Quran 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 nally 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 fteenth 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,

Religious Language Management 75 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 developed 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 IndoEuropean 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 of cial 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 of cial 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 peacefully, spreading particularly in the island regions like Indonesia while the mainland 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 rst communities were established (Shackle, 2001). In many areas, Arabic persisted for ritual inculcation of a more or less mechanical recognition of Arabic suf cient for recitation of the Quran (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 religious 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 Quran 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 rst 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 Quran into the vernacular (the translations into mediaeval Persian and Ottoman Turkish were

76 Bernard Spolsky exceptions) and required the teaching of Arabic. Schoolchildren in Quran 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, continues 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 unacceptably 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 Quran and the corrupt spoken language. The higher status afforded to the written language led to diglossia (Ferguson, 1959; Hudson, 2002). Quranic literacy bene ted 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 fundamental commitment to the language of the Quran. 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 religious matters.

4.5 Other Religious Language Management

Hinduism from the religious point of view is better considered as a variety of religious 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 de nes 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 communities and authoritative interpreters that were essential for ef cient administration of colonial India. It was essentially the elitist Brahman forms and traditions that

Religious Language Management 77 were stressed. The multiplicity of Indian religious tradition was thus standardized and uni ed, in a form consistent with Western models. This development paralleled 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 phonetics, 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 differentiating feature, and all languages were considered equally able to express the content. While the Classical languages (Sanskrit, Arabic and Pali) remain connected 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 Quran.

78 Bernard Spolsky Translating these generalizations to any speci c 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 divinity) 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 translate the sacred text into a language that the congregants do know. Also, as a general rule, the ministers 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 Germanspeaking 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 written in a language they do not understand: prayer books may then include translations 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 in uence 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 of cials 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 continued 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 modi ed by the efforts of some religions to convert others. At the same time, globalization and the global

Religious Language Management 79 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 eld of religious language management.

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 signi cant 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 Rushdies 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 of cial in the West Bank and Gaza and alongside Hebrew in Israel.

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 in uenced 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), 4767. Birch, B. M. (1995), Quaker plain speech: A policy of linguistic divergence. International Journal of the Sociology of Language, 116, 3959. 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. 126127. Collins, P. (2003), Storying self and others: The construction of narrative identity. Language and Politics, 2, (2), 243264. Fabian, J. (1983), Missions and the colonisation of African languages: Developments in the former Belgian Congo. Canadian Journal of African Studies, 17, (2), 165187. Ferguson, C. A. (1959), Diglossia, Word, 15, 325340. 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. 253265. Fishman, J. A. (1972), Domains and the relationship between micro- and macrosociolinguistics, in J. J. Gumperz and D. Hymes (eds), Directions in Sociolinguistics. New York: Holt Rinehart and Winston, pp. 435453. 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 Paci c. Glinert, L. (1991), Language choice and the Halakhic speech act, in R. L. Cooper and B. Spolsky (eds), The In uence of Language on Culture and Thought: Essays in Honor of Joshua A. Fishmans Sixty-Fifth Birthday. Berlin: Mouton de Gruyter, (pp. 157182). Gogonas, N. (2007), Ethnolinguistic Vitality and Language Maintenance in Second-Generation 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. 8384. Harlow, R. (2007), Maori: A Linguistic Introduction. Cambridge: Cambridge University Press. Hudson, A. (2002), Outline of a theory of diglossia. International Journal of the Sociology of Language, 157, 149. 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. 8596. Isaacs, M. (1999), Contentious partners: Yiddish and Hebrew in Haredi Israel, International Journal of the Sociology of Language, 138, 101121. 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. 165179.

Religious Language Management 81

Killingley, D. D. (2001), Hinduism, 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. 5254). King, R. (1999), Orientalism and the modern myth of Hinduism, Numen, 46, (2), 146185. 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, Shannon, Singapore, Tokyo: Elsevier, pp. 6566. 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 (eds), Concise Encyclopedia of Language and Religion. Amsterdam, New York, Oxford, Shannon, Singapore, Tokyo: Elsevier, pp. 6062. 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. 141164). Peters, F. E. (2003), The Monotheists: Jews, Christians and Muslims in Con ict and Competition. Princeton, NJ: Princeton University Press. Pickering, W. S. F. (2001), Blasphemy, 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, p. 240. Rabin, C. (1981), What constitutes a Jewish language? International Journal of the Sociology of Language, 30, 1928. 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 Languages and Linguistics. Cambridge: Elsevier. Sawyer, J. F. A. (2001), Christianity in Europe, in J. F. A. Sawyer and J. M. Y. Simpson (eds), Concise Encyclopedia of Language and Religion. Amsterdam, New York, Oxford, Shannon, Singapore, Tokyo: Elsevier, pp. 3335. Shackle, C. (2001), Islam in South Asia, in J. F. A. Sawyer and J. M. Y. Simpson (eds), Concise Encyclopedia of Language and Religion. Amsterdam, New York, Oxford, Shannon, Singapore, Tokyo: Elsevier, pp. 6265. 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. 119169.

82 Bernard Spolsky
Spolsky, B. (1983), Triglossia and literacy in Jewish Palestine of the First Century. International Journal of the Sociology of Language (42), 95110. Spolsky, Bernard. (1991a). Control and democratization of sacred literacy. In Samuel Rodin (Ed.), Encounters with Judaism: Jewish Studies in a non-Jewish World (pp. 3753). Hamilton: Waikato University and Colcom Press. Spolsky, B. (1991b), The Samoan language in the New Zealand educational context. Vox, 5, 3136. Spolsky, B. (2002), Prospects for the survival of the Navajo language: A reconsideration. Anthropology & Education Quarterly, 33(2), 124. Spolsky, B. (2003a), Reassessing Maori regeneration. Language in Society, 32(4), 553578. Spolsky, B. (2003b), Religion as a site of language contact. Annual Review of Applied Linguistics, 23, 8194. Spolsky, B. (2005), Maori lost and regained in A. Bell, R. Harlow and D. Starks (eds), Languages of New Zealand (pp. 6785). Wellington: Victoria University Press. Spolsky, E. (2007), Word vs. Image: Cognitive Hunger in Shakespeares 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 Multilingual and Multicultural Development, 4, (6), 459470. 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. 326336. 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. 2538. Weeks, T. R. (2002), Religion and Russi cation: Russian language in the Catholic Churches of the Northwest Provinces after 1863. Kritika: Explorations in Russian and Eurasian History, 2(1), 87110. 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. 459470. 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, 7180.


5 Language and Culture1

N. J. En eld

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 (En eld 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 inhabiting positions in a vast network of social statuses (Linton, 1936; Kockelman, 2006b). Attempts to de ne 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 representations, 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. En eld

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 public 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 language 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 isnt the case. Both language and culture are cognitive phenomena, which means that they are grounded in a capacity for exibly solving means-ends problems, and a capacity for using mental representations displaced from an immediate context. This is captured in the de nition 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 exibly, with the individual 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 de nitive components of cognition, the latter, mental representation, has been dominant in research on language. The classic Saussurean sign signi er standing for signi ed 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 re exive 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 de nitional 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


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 structures; Norman, 1991; C. Goodwin, 1994, 2006; Hutchins, 1995, 2006; En eld, 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 livelihood 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 de ne 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 capacity 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 speci c, token relationship (e.g. between me and my brother Matt) as well as a type of

86 N. J. En eld relationship (e.g. between a man and his brother). Types of relationship in turn de ne types of social statuses and identities which will be de ning elements of higher-level social structure (Linton, 1936: 113131, Radcliffe-Brown, 1952; Lvi-Strauss, 1953; Nadel, 1957; Hinde, 1976; Sacks, 1992; Dunbar and Spoors, 1995; Hill and Dunbar, 2003; Pomerantz and Mandelbaum, 2005; En eld 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, compassion and common identity (Henrich et al., 2004; Boyd and Richerson, 2006). On the other hand are the Machiavellian instincts that license competition, deception, 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 interested in marking boundaries and establishing social difference (M. H. Goodwin, 1990: 141225, 2006; cf. Goffman, 1959, 1967). However, these ostensibly contrastive 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 sel sh (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 relationships 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) relationships enacted in interaction co-de ne the roles and identities that will ultimately de ne 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 de ned by how they each stand towards some common reference point (with associated de nitive commitments and entitlements). For instance, if A and B are both members of a local cricket club, this is a potential basis for de ning a relationship between them through external grounding. Such relations may be negatively de ned 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

Language and Culture


external ground. By contrast, if A and B stand in a reciprocally grounded relationship, then the rights and responsibilities associated with A and Bs incumbency in that relationship are mutually de ned: e.g., if A is Bs father, B is necessarily As 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 individualism 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 relationships, 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 individual 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 speci c context, and in a particular kind of social world, as de ned, in part, by species-speci c determinants such as pro-social instincts, social intelligence capacities and structural 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. En eld 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 different 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; En eld and Levinson, 2006, inter alia):5 perspective-taking (awareness of others perceptual states) false belief understanding (truth vs. peoples representations of it) pro-social instincts (altruism, group living, ethnic co-membership) cooperative instincts (capacity for exible joint action toward a mutual goal) Machiavellian instincts (dominance, coalition-building, manipulation, ethnic distinction) intention-recognition (attribution of knowledge, belief, desires) an intentional stance (intention-attribution to the non-mental realm) management and exploitation of mutual knowledge (Schelling thinking) a uid symbolic capacity (sensitivity to social convention) docile cultural instincts (propensity to adopt the norms of ones group) socially anchored emotional and moral instincts (motives to adhere to and regulate social norms) These are presumed to re ect universal human capacities, de nitive of the cognitive 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 in ection 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 management (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 of cial 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 environment is by brute force wielded upon physical objects (say, chopping wood for re). 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

Language and Culture


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 (En eld, 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; BaronCohen, 1995; En eld 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; Gar nkel, 1967; Brandom, 1979; Kockelman, 2006b). Social agency is built on this kind of normregulated semiotic commitment, de ned 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 interpretants (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 (En eld, 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 speci c contexts. This is the basic insight behind Grices work on meaning (1957) and conversational inference (1975) (cf. also Goffmans idea of framing; 1974, 1981). Grices 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 (Grices implicature) is a smart solution to a thorny bottleneck problem in human communication: we think fast but we speak slow. While Grices (or Sperber and Wilsons, or Levinsons) 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. En eld culturally are the local premises (norms) which feed the process of inference, not the inferential processes themselves (cf. En eld, 2002, Evans, 2003).

5.7 Language and Culture in Problem-Solving

From this perspective, we can characterize language and culture as tools and contexts for communicative action in management of the social world. The domain of pragmatics, where language and culture come together, may be de ned as social problem-solving. Consider what is meant by problem-solving in a more general sense.7 If I dont eat every day I get hungry. To solve this, I might engage in a complex 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-speci c facts. Notice how problems and solutions are nested. Once I have committed to a certain solution, this raises new imperatives (cf. Dunbar, 1988: 2628). For example, as a professional 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 problemin-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 semiotic 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 Af liation

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 af liational imperative (En eld 2006).10 The informational imperative is to ensure

Language and Culture


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 suf cient for current communicative purposes. At the same time, the af liational 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 relationship such that it may be evaluated by participants and onlookers (En eld, 2006). A relationship thinking approach puts this in the foreground. We might also refer to the af liational imperative as micro-political or coalitional, 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 speci c 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; En eld, 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 af liation and coalition.

5.9 Conclusion
The theme of this chapter has been that language and culture are deeply implicated 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 de ne 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-de ning and overlapping. 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; En eld, 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.

1 This chapter incorporates revised sections of an article titled Relationship thinking and human pragmatics, published by Elsevier in Journal of Pragmatics. I am

92 N. J. En eld
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. Any investigation of language or culture has to cope with their re exive 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 property 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 meaning 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. Unlike Ingold I do not see this as incompatible with population thinking (a concept attributed to Darwin: Mayr, 1964: xixxx; 1970; 1982: 4547; see Hinde, 1991: 585586). This does not apply to other complex societies such as those of the ants, since they are not individualized in de Waal and Tyacks sense. These are not necessarily qualitatively distinct. The list merely represents a range of different angles on social intelligence from a range of disciplinary traditions. These are also applied in interpretations of the non-social world (Lvi-Strauss, 1966; Goody, 1995; Atran, 2002). I will sometimes distinguish between imperatives as problems which demand solutions, and strategies as the particular solutions chosen (Dunbar, 1988). See Schutz (1970) and predecessors for a distinction between the because motives which focus on the states of affairs which give rise to actions (Im picking berries because Im 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 (Im picking berries in order to eat them). Among the set of problems-in-need-of-solution, some will be generically present across cultures (e.g. the need to deal with signi cant problems of speaking, hearing 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 speci ed 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. Its 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. These are akin to Goffmans system versus ritual constraints in face-to-face interaction (Goffman, 1981). Paul Kockelman (personal communication) points out that these correspond roughly to Jakobsons referential and phatic functions of language, two among his six general functions (the other four being emotive, poetic, conative and metalingual; Jakobson, 1960).

4 5 6 7 8


Language and Culture


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 cognitive cross-species comparison, Journal of Comparative Psychology, 121, (227240). Boyd, R. and Richerson, P. J. (2006), Culture and the evolution of the human social instincts, in N. J. En eld and Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition, and Interaction. London: Berg, pp. 453477. Brandom, R. B. (1979), Freedom and constraint by norms. American Philosophical Quarterly, 16, (3), 187196. 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: Cambridge University Press. 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. En eld (ed.) Ethnosyntax: Explorations in Grammar and Culture. Oxford: Oxford University Press, pp. 3151. 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, 273290. En eld, N. J. (2000), On Linguocentrism, in Explorations in Linguistic Relativity. Martin Ptz and Marjolijn H. Verspoor (eds), Amsterdam/Philadelphia: John Benjamins, pp. 12557. En eld, N. J. (ed.) (2002), Ethnosyntax: Explorations in Grammar and Culture. Oxford: Oxford University Press. En eld, N. J. (2003), Linguistic Epidemiology: Semantics and Grammar of Language Contact in Mainland Southeast Asia. London: RoutledgeCurzon. En eld, N. J. (2005), Areal Linguistics and Mainland Southeast Asia. Annual Review of Anthropology, 34, 181206.

94 N. J. En eld
En eld, N. J. (2006), Social consequences of common ground, in N. J. En eld and Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition, and Interaction. Oxford: Berg, pp. 399430. En eld, N. J., (2007), Meanings of the unmarked: How default person reference does more than just refer, in N. J. En eld and Tanya Stivers (eds), Person Reference in Interaction: Linguistic, Cultural, and Social Perspectives. Cambridge: Cambridge University Press, pp. 97120. En eld, N. J. (2009), The Anatomy of Meaning: Speech, Gesture, and Composite Utterances. Cambridge University Press: Cambridge. En eld, N. J. and Levinson, S. C. (eds) (2006), Roots of Human Sociality: Culture, Cognition, and Interaction. London: Berg. En eld, 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, 1340. 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. Gar nkel, 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), 606633. Goodwin, C. (2006), Human sociality as mutual orientation in a rich interactive environment: Multimodal utterances and pointing in aphasia, in: N. J. En eld and Stephen C. Levinson (eds), Roots of Human Sociality: Culture, Cognition, and Interaction. London: Berg, pp. 97125. 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 University Press. Granovetter, M. (1973), The Strength of Weak Ties, American Journal of Sociology, 78, 13601380. Granovetter, M. (1978), Threshold models of collective behaviour, American Journal of Sociology, 83, 14201443.

Language and Culture


Grice, H. P. (1957), Meaning, Philosophical Review, 67, 377388. (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. 4158. (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, 1538. Hill, J. H. and Bruce M. (1992), Language and world view, Annual Review of Anthropology, 21,381406. Hill, R. A. and Dunbar, R. I. M. (2003), Social network size in humans, Human Nature, 14, 5372. Hinde, R. A. (1976), Interactions, relationships, and social structure, Man (New Series), 11, (1), 117. 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 Anthropological Institute, 26, (4), 583608. 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. 303321. Hutchins, E. (1995), Cognition in the Wild. Cambridge, MA: MIT Press. Hutchins, E. (2006), The distributed cognition perspective on human interaction, in N. J. En eld and S. C. Levinson (eds), Roots of Human Sociality: Culture, Cognition and Interaction. Oxford: Berg, pp. 375398. 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, 208229. 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. 350377.

96 N. J. En eld
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. 130147. Jolly, A. (1966), Lemur social behavior and primate intelligence, Science 153, 501506. Kockelman, P. (2005), The semiotic stance, Semiotica 157, 233304. Kockelman, P. (2006a), Representations of the world: Memories, perceptions, beliefs, intentions, and plans, Semiotica, 162, (14), 73125. Kockelman, P. (2006b), Residence in the world: Affordances, instruments, actions, roles, and identities, Semiotica 162, 1971. Kockelman, P. (2007), Agency: The relation between meaning, power, and knowledge, Current Anthropology, 48, (3), 375401. 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. 380405. Kuper, A. (1999), Culture: The Anthropologists 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 prerequisites, 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. Lvi-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-CenturyCrofts. 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), 682707. Norman, D. A. (1991), Cognitive artifacts, in J. M. Carroll (ed.) Designing Interaction: Psychology at the Human-Computer Interface. Cambridge: Cambridge University Press, 1738. Owings, D. H. and Morton, E. S. (1998), Animal Vocal Communication: A New Approach. Cambridge: Cambridge University Press.

Language and Culture


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. 149171. 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. En eld and S. C. Levinson (eds), Roots of Human Sociality: Culture, Cognition, and Interaction. Oxford: Berg, pp. 7096. 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: Cambridge University Press. Sidnell, J. (2005), Talk and Practical Epistemology: The Social Life of Knowledge in a Caribbean 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. 1155. 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), 664670. 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 pragmatics, Language 63, (1), 95114. 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.


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 discourse (ABSD), and Web-borne situated discourse (WBSD), with the term fourborne 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 discourse. Following Chilton (2004: 16) I shall use languageL to refer to the human capacity for language (cf. I-language in Chomsky, 1986: 22), languagel to refer to speci c languages such as Chinese, English, French, etc. The use of languagel/u will be discourse, i.e. Chomskys E-language. Nothing will be said about languageL. The focus will be on the interaction between languagel and discourse. Note that the notion of language to be dealt with hereafter will be by default languagel unless it is indicated otherwise. Furthermore, the Chinese language is the object of investigation, and the Chinese social, political and cultural context is taken for granted. As is well known, theories are abundant about languagel and discourse. It obviously 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 ow 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 Wittgensteins three metaphors about language are reviewed and examined (section 6.3). They are important because they serve as the conceptual foundation for bridging language and the four-borne discourses. Section 6.4 makes a demographic analysis of the four-borne discourses, followed 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


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 uniform houses. In view of Chinese, the Chinese language city arguably started being built as early as Peking Man (Homo erectus, 400,000200,000 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 (17651122 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 telecommunications technology rst in telephone calls, then radio and TV broadcasting 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 situated 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 6.1a6.1c show three screen shots visually illustrating the three-borne situated 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 Wittgensteins Three Metaphors
As Aitchison (2003: 39) observes, the usefulness of metaphor in theory building has of course long been recognized, in a variety of scienti c elds. 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: 4344). 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 Saussures game is primarily a chess game, Wittgensteins notion of game is a series of games with family likeness, ranging from chess game to games such as ring-a-ring-a-roses (see Wittgenstein, 1997 [1953]: 5e). Moreover, Wittgensteins 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 Wittgensteins use of game metaphor goes beyond highlighting a games 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 activity or of a form life that Wittgensteins game metaphor is related to another metaphor of his, viz. language as an ancient city. In Aitchisons review, Wittgensteins city metaphor gets lost in what she calls buildings metaphor. Under the subheading 5.7 Buildings (italics original, 2003: 45),

Four-Borne Discourses 101 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), Wittgenstein 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 Fillmores likening language to a house (see Aitchison, 2003: 45). Besides the game and city metaphors, 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 Wittgensteins 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 integrated picture of language conceived by Wittgenstein. First, the language-as-anancient-city metaphor is not accidental, but rather plays a major role in Wittgensteins philosophizing in general. He starts his Philosophical Investigations with what he calls primitive language, the term being just brought in without clearly de ning it. It has nothing to do with the term as used in the literature of early anthropological linguistics. Wittgensteins 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 Philosophical 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 Wittgensteins 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 metaphor 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 Wittgensteins 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 languagel). It has to be pointed out here that Ayers (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 rst speaker to make the rst use of his or her private

102 Gu Yueguo language before the rst use becomes adopted by others to make the rst public language-game. Unlike Wittgensteins primitive language which is constructed and deliberately simpli ed, Peking Mans 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 reference, 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 existence mediated by telecommunications technology, i.e. telephone, radio and traditional 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 technology, 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 Webborne 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 others resources. For instance, CCTV (news channel) has recently run a program called Meiti Guangchang ( ), the content of which is to televise newspaper cuttings from some in uential newspapers. The cityscape of the Chinese language thus is not a static stack of discourses, but a dynamic ux with discourses owing around.

Four-Borne Discourses 103 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 fourborne 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 Peoples 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 population 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 partially 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 absolute number for the rst 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 internet 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


ta to

ul op lp

io at I

n e

at er llit W

ul ad (b

ts le ab to

) ad re





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 xed 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 presentday 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 individuals 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

Four-Borne Discourses 105 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 ones 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 con ned 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 sur ng 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 nally (7) fourborne 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 ( ), a well-known TV presenter). It can be estimated (almost impossible to pinpoint accurately) that the last category, viz. the four-borne idiolect speakers are the smallest minority, while the majority is the third category, viz. expanded one-borne idiolect speakers. So the seven idiolect categories make a arrowhead-shaped gure, as shown in Figure 6.3. (Note the gure is more metaphorical than statistically accurate. No one can work out exact statistic gures. 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 fourborne 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) accessibility 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 categories Receptive and productive types
LBSD WWBD WBSD ABSD LBSD WWBD WBSD ABSD (receptive only) LBSD WWBD WBSD (receptive only) ABSD (receptive only) LBSD WWBD ABSD (receptive only) LBSD WWBD (receptive only) ABSD (receptive only)

Speaker types

4-borne idiolects

literary or non-literary writers, Bloggers, and TV presenters, e.g. Zhao Zhongxiang

3-borne idiolects

literary or non-literary writers as well as Bloggers, Website keepers

expanded 2-borne idiolects

literary or non-literary writers, editors, academics

2-borne idiolects

literates with formal higher level education and producing practical writings

expanded 1-borne idiolects

enriched 1-borne idiolects

literates with some formal literacy education

illiterates having access to ABSD TV programs in local dialects or in Putonghua


pure 1-borne idiolects


illiterates in the remote mountainous areas

Figure 6.3 An arrowhead-shaped distribution pattern of idiolects.

6.5.1 Analytic Parameters Situated discourse The term situated discourse as type name is used in contrast with non-situated discourse. It is an abstraction of all those instances of language use that have the ensuing characteristics. The language use is 1. situated to a particular spatial and temporal setting of a social situation (as de ned 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.

Four-Borne Discourses 107 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 political boundaries or the like (Random House Websters 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, peoples 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 physically 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 (Hgerstrand, 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. Hgerstrand, 1975 and 1978). Nowadays people are of course much better off than people in the primary orality 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 identi ed, 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. Human space-time movement and activity zone Thanks to the LBSDs features discussed above, participants movement in physical 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 con guration of the web is profoundly structured by the life patterns of a speech community and the individual language users daily activity zone (see Gu, 2002b: 144145 for a case study of week-long activities of Mr X). In the authors eld study in Inner Mongolia, for instance, it is found that there are local herdsmen whose daily activity 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 consequences 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 generation becomes bilingual, i.e. in Putonghua and the Spoken Mongolian. Social Space-Time The previous section has shown the coupling of human bodily movement in physical space and time with individual members everyday activities. This coupling provides 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 spacetime). The concept can be demonstrated with the age-old household registration practice, the earliest written record of which was found in West Zhou Dynasty (1121771 BC).3 In Ming Dynasty (13681644 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 certi cate (luyin ). 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- rst 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 SpaceTime, since they cannot become registered as regular household members in their host cities, thus being denied access to the bene ts 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 grandparentsgenerations 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 countys 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 grandparentsgenerations LBSD (i.e. the Spoken Mongolian) with oral comprehension failing fast on its way.

Four-Borne Discourses 109 Natural multimodality and total saturated experience In the LBSD, the face-to-face, co-present interaction between participants is naturally 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 experienced by participants in esh 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). 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 prominent 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 primarily facilitated through human memory in the sense that the LBSDs past occurrences, 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 Barkers sense, Barker, 1968), e.g. a behaviour setting xed 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 below; see also Gu, 2008 for further discussion). Accessibility and availability Accessibility is associated with the affordance the four-borne discourses are intrinsically able to provide, whereas availability refers to the chances individual members 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 The WWBD Situatedness of discourse: The WWBD is a prototypical case of non-situated discourse, using those features of situatedness discussed in and 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 writers unique identity can be hidden from a readers knowledge. Human spatial-temporal movement : In the WWBD, the writers original space-time is embodied symbolically in the WWBD. It is always separate from the readers spacetime. 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 costeffective. Having said this, thanks to the separation and discontinuation of spacetime, 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 de ned in Section is largely constructed, 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, 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-modalized 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, prosodies, 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 Words 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 accumulating knowledge and cultural heritage. According to Bodde (1957: 10), it has

Four-Borne Discourses 111 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. The ABSD Situatedness of discourse : While the LBSD is always con ned to a xed location such as a building, an of ce, 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 situatedness: TV series, lms and other imaginative multimedia programmes are no longer situated to a particular behaviour setting, or to a here-and-now gathering of copresent 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 physical space (e.g. the projecting TV studio or where the body-person stays), and the air space de ned 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. (Thats why Osama Bin Ladens TV broadcasts needs to be veri ed!). One of the greatest advantages of the ABSD is that the projected persons 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 Chinas CCTV plays is to disseminate what has been xed 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 ABSDs role is drastically different. ABSD producers and the projected persons gain a considerable 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. analogue cassettes, digital hard disks, etc. The great advantage of this technologymediated 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 ABSDs oral-aural and visual modality, being close to the LBSDs 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). The WBSD Situatedness of discourse: In the LBSD, each participant (i.e. body-person in one) is uniquely identi ed, and veri ed 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 netconferencing, 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 de ned 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

Four-Borne Discourses 113 ABSD, the projected persons movement over the projected space (i.e. the travelling 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 contrast, space and time are merged into each other. The projected persons/messages y 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 constructive 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 de ned in 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 inde nitely. 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 knowledge and cultural heritage. Accessibility and availability: The current Web technology, due to the limitations of HMI discussed above, con nes its accessibility to literates. Its availability is also limited 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 brie y return to the relation between languagel and discourse, or in Wittgensteins metaphorical terminology, 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 xed 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 xed 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 individuals are engaged in social mobility, and migration. This dynamic ux is counterbalanced 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 interpersonal 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, reproduction 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 ABSDs contribution to the reproduction of language use lies in its capacity for mass publicity and awareness-raising. The ABSDs 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 orthographic text into the ABSD. Time will come when the TV series writers compose them straightway without composing in the WWBD rst, like painters just draw without putting what is in mind rst 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 wont 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 rede ne the traditional status of orality and literacy, which is currently 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

Four-Borne Discourses 115 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 inspirations primarily from the ensuing areas, discourse analysis, corpus linguistics, sociolinguistics, 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 fragmented, leaving many loopholes and undefended assumptions that are only being xed in other places. Having said this, the chapter still has a focus of its own: It presents a conceptualization of Wittgensteins 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 Wittgensteins 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 language 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 xed once for all, but evolving, as time goes by (note that historicity is an essential ingredient in Wittgensteins 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 multimodalization (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 Hgerstrands theory of time geography in particular. The review of Hgerstrand 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 internet with a great amount of literature having been accumulated. As a space-saving

116 Gu Yueguo strategy, Danesis (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, telecommunications 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. Media and a Fundamental Metaphor Theorization in media studies is underpinned by a fundamental metaphor conceived by McLuhan, one of the founding gures 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 genuinely invented by man that have been scrutinized in media studies: print, telecommunications and digital computing. Semiotic approach to media studies McLuhans (1964) work reads more like a learned lyric than an academic thesis. He is regarded by some as the erudite but generally unintelligible Toronto professor (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 in uential 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 heritage. 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 dimensions (personal, social, historical). Figuring out what the meanings of Y are constitutes the sum and substance of semiotic method. This procedure is generally referred to as interpretation.

Four-Borne Discourses 117 Adopting Saussures terminology, X is the signi er; the meaning or meanings, Y, that it generates is the signi ed.

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, McLuhans conceptualization of the medias extensions of man departs from the abstract state of man without language, while our departure point is men living a life with speech in a physical space-time setting i.e. the LBSD. McLuhans 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 mans physical and mental capabilities. In our conceptualization, men living a life with speech is not regarded as being mediated, or rather as being naturally mediated, whereas in McLuhans, 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 re ect our different conceptualization. The LBSD is naturally multimodal 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 produced 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 reference framework for studying technologically mediated discourses. The methodology adopted in studying the process of the transformative partnerships 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 perspective 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 rst is a critical strand. It examines media discourse critically (e.g. van Dijk, 1988; Fowler, 1991) by bringing 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. Scollons notion of mediated discourse is derived from Wertschs 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 Bergsons 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 Wittgensteins terminology. 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.

Four-Borne Discourses 119 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 Ongs 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- owing channels for the previous two. It is by all means possible that new yet-to-be-borne technologies will come and add more layers 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 un nished 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)!

1 Source: www.news.cn ( ) 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 see 4 , see , 2002 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.

Aitchison, J. (2003), Metaphors, models and language change, in H. Raymond (ed.), Motives for Language Change. Cambridge: Cambridge University Press, pp. 3953. Argyle, M., Furnham, A. and Graham, J. A. (1981), Social Situations. Cambridge: Cambridge University Press.

120 Gu Yueguo
Ayer, A. J., (1968), Can there be a private language?, in G. Pitcher (ed.), Wittgenstein: The Philosophical Investigations. London: Macmillan, pp. 251266. Barker, R. G. (1968), Ecological Psychology: Concepts and Methods for Studying the Environment of Human Behavior. Stanford: Stanford University Press. Bodde, D. (1957), Chinas 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, ( ) 21 in Globalization and the 21st Century, Beijing: Social Sciences Publisher, pp. 484500. 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. 137185. Gu, Yueguo, (2006a), Agent-oriented modeling language, Part 1: Modeling dynamic behavior (AML) 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, 262, pp. 127167. 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. Hgerstrand, T. (1978), A note on the quality of life-times, in Carlstein, Parkes and Thrift (eds), pp. 214224. Hgerstrand, T. (1975), Space, time and human conditions, in A. Karlqvist (ed.), Dynamic Allocation of Urban Space. Farnborough: Saxon House, pp. 312. 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.

Four-Borne Discourses 121

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. Cambridge: Cambridge University Press. Wittgenstein, L. (1997[1953]), Philosophical Investigations. 2nd edition. Translated by G. E. M. Anscombe. Oxford: Blackwell Publishers.


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, organization/network theory and social constructivism, researchers have sketched the macro-frames which in uence 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 understand communication in the modern organization from different viewpoints and with a focus on different problem areas. The theoretical and methodological perspectives 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 professional discourse have been in uenced by a series of changes taking place in recent decades. Technological advances have facilitated a globalization of working life, while lifelong learning, exibility, mobility and diversity have come to be key values in the global economy. In organizations throughout the world, we nd a widespread 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 market. Throughout the world, we nd 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 workplace diversity which characterize a globalized economy also raise issues of multilingualism 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 discourserelated problems facing managers and employees in the globalized economy. In the various sections of the chapter I discuss these problems from different viewpoints. 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


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 discourse from a workplace perspective to distinguish two types of multilingual workplaces: 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 correspondence between the organizational structure and discourse. This correspondence is seen here as a two-way relationship. Creating texts (written, spoken, and computer-mediated) therefore forms part, and an important part, of an organizations work. Texts1 are in uenced by the social context, thus re ecting 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 rst study I discuss provides an example of the simple sociolinguistic order of discourse, which we nd in close-knit workplace groups. Gunnarsson (1997) presents a study of text production in a local government of ce 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 ow of texts, i.e. all texts emanating from the of ce during a xed 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 re ected in the communicative structure: the hierarchical social structure was re ected in communicative patterns relating to in uence and supervision; the informal group formation was re ected in informal joint writing efforts. The social organization at the of ce was re ected 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 nding 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 of ces. This group identity manifested itself in many ways, also in relation to writing. The interviewees made it plain that in this of ce 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 of cial idiom. One conclusion that can be drawn from this is that even small groups, such as a local government of ce with 35 employees, consider that they have the right to be innovative 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 re ected 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 complex and multi-dimensional picture of the correspondence between organizational structure and discourse. Large organizations function within various social frameworks 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 Communities2, which studied texts and text production in four mainly national writing communities banks, structural engineering rms, 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 sociolinguistic, which involved studying the relationship between national culture, organization and texts. The study comprised interviews, collecting texts, corpus construction and textual analysis. Altogether, 70 executives and employees, responsible for or involved in information activities and the production of texts, were interviewed. The team also collected samples of the types of texts produced in the environment 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 argumentation, discourse markers and images and image creation in companies. The results of the analyses of the interviews and the collected texts reveal interesting differences between the sectors between banks and structural engineering

Discourse in Organizations and Workplaces


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 nding of interest is the homogeneity within each organization. A fairly homogeneous picture could be drawn of each company, 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 at 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 intertwined with various levels of other orders leading to a multifaceted and multilayered disorder.

7.2 The Construction and Maintenance of an Organizational Self

From the point of view of the top management of large national and international 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 important 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 existing organizational culture.

126 Britt-Louise Gunnarsson 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 unintelligible and insigni cant. 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 FinnishSwedish 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 in uenced 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, con icts, individual attitudes and perspectives contributed to the successive distortion of the top managements visions. Johanssons 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 externally 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 IKEAs trademark, and it has obviously managed to turn Swedishness into an image that sells. The companys products are given Swedish personal names, like Bosse, Bjrn 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 , and are therefore not avoided, but in fact deliberately used to strengthen IKEAs 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.

Discourse in Organizations and Workplaces


As part of its Swedish image, in other words, IKEA wishes to introduce the more egalitarian 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 Jmtelid (2001, 2002). The concept of parallel writing is used for the analysis of the multilingual production of text in the Electrolux corporate group, where English is the of cial language (cf. Gunnarsson and Jmtelid, 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. brochures for vacuum cleaners, a textual base in English is devised at the main of ce in Stockholm, Sweden, and circulated to the sales companies in the other countries. The textual base then provides the raw material for the various consumer brochures produced in different languages and intended for different cultures. As Jmtelids analyses show, the different sales companies have chosen to incorporate different ideas from the textual base circulated by head of ce. She also found marked differences in the arguments for the product presented in the brochures, the choice of illustrations and the styles used. There were thus clear differences 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 (Jmtelid, 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 important difference between texts on a companys 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 speci ed reader of the printed document is replaced by a potentially manifold readership: shareholders, staff, journalists, politicians, former and prospective customers in countries all over the world can view the company website. Although the company 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 companys 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 companys 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

128 Britt-Louise Gunnarsson companys 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 ve transnational companies: ABB, Astra Zeneca, Electrolux, Ericsson and Scania. All these companies claim to be world leaders in their elds: 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 worlds leading pharmaceutical companies, Electrolux is the worlds 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 worlds 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.comsites. All the companies have a .com-site in English, functioning as a start page. For all ve companies, we also nd the option of selecting a country on the start page. The number of country links varies among the ve companies, which of course re ects 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 practices of the country websites. In which language(s) do the ve 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 ve 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 of cial 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 Ericssons country websites. All the ABB websites use the main national language, but for ve 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 nd the national language(s). On two of Scanias 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.

Discourse in Organizations and Workplaces


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 gures 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 ve companies studied, accessibility seems to be reserved for those prospective 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 todays 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 discuss 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 lingua franca of these organizations. In countries where English is not the mother tongue, this creates different communicative problems, in relation to both outgoing 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

130 Britt-Louise Gunnarsson 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 FinnishSwedish 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 communication between buyers and sellers in which foreign languages have been traditionally 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 writers, 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 ve 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 quali cations 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 rst 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. ABBs 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 ve 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 ve transnational companies multilingualism means bilingualism, i.e. a knowledge of English and Swedish. Only 10 per cent of the advertisements mention languages other than Swedish and English as a quali cation. English, on the other hand, is explicitly mentioned in 70 per cent of the advertisements. Swedish gures less prominently as a quali cation, and is explicitly required in only 34 per cent of the adverts.

Discourse in Organizations and Workplaces


Of course, the variation in terms of the quali cations 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 Jmtelid 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 widespread 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 presentation brochure obviously has to be correct, it still has to be possible to send a message 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 dif cult words, and that people shouldnt 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).

132 Britt-Louise Gunnarsson Other studies have been more negative about the levelling of cultural differences in international meetings. Fant (1992) on the basis of his analyses of crosscultural 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 rst 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 Brestam (2005), who has analysed meetings within Nordea, another Nordic company with English as its of cial 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 differences which for instance relate to the overall structure of the meeting discourse tends to favour a more Americanized style (71). Fants and Brestams 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 of cial documents are written in English. Language knowledge and language de ciencies 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 of ce in Sweden. She recorded and analysed the information ow 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 contents 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. Johanssons study shows, among other things, that the use of English for important internal documents might increase the distance between senior management and ordinary employees.

Discourse in Organizations and Workplaces


In many cases, the divide between those with mastery of English and those without increases the gap between native and non-native speakers. The choice of English 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 rst 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 bilingualism is thus gradually developing in Sweden among young people and educated 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 rst 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 rst, when moving to Sweden, is of course also Swedish and not English. In organizations which use English as their of cial 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 of ce 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 of ce 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 oor, 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 language is necessary not only for advancement and a career but also for social integration into the working group. Humour, jokes, stories, anecdotes form part of the workplace discourse and contribute to the establishment of friendship and collegiality at work and also to the avoidance of unnecessary con icts. Social workplace patterns related to friendship, power and dominance often re ect 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

134 Britt-Louise Gunnarsson 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 immigrants 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 situations 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 collaboration 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 individuals 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 represent 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 ve 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 companies, 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 languages, 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 AstraZenecas web pages, one is an immigrant to Sweden, born in Lebanon, and of the 14 featured on Scanias site, one is an immigrant to Sweden, born in Iran, and one a visiting student from Germany. Both these companies could be said to give a fairly good picture of the current Swedish job market. As regards Scania, it should be noted that the German masters degree student is said to be uent in Swedish. The companys monolingual Swedish culture is not therefore affected by the presence of this German girl. Though all ve companies explicitly stress various societal values, e.g. diversity, on their websites, the analysis of the success stories on their career-oriented web

Discourse in Organizations and Workplaces


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 in uenced by a series of changes taking place in recent decades. Technological advances have coincided with a globalization of working life and lifelong learning, exibility, 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 nd 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 signi cance. 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 in uenced by the technological advances10, and phone calls and videoconferencing 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 phenomenon, its function and impact on discourse. Text, speech, graphs, recorded sound and movies are interwoven in todays communication in a way which was not previously 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.

136 Britt-Louise Gunnarsson Though these studies offer glimpses of the interlacement of discourse and technology 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 exible 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 middle 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 organizations must nd 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 communicative skills. Organizations competing on the global market need to be exible 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 exible person who is continuously learning as she/he moves within the organization and between organizations, thus taking on new jobs and performing new tasks. Instead of unskilled workers performing routine tasks, organizations need staff who can act independently, 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 exibility and responsibility from individual employees. Organizational structures based on workforce mobility and workplace diversity also make higher demands of the language 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 in uenced 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 dif culties could be overcome. In this last section I dwell on a few topics which relate to the various issues discussed earlier in this chapter.

Discourse in Organizations and Workplaces


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 af liations 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 workplace group? Fewer levels and reduced hierarchies result in vagueness and tension for individual employees. The need for workforce exibility also leads to a form of growing 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 at organizations. Another topic which would be worth looking into relates to the role of discourse for bottom-up in uence. 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 in uence 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 management level? How does top management try to steer the shop oor workers? What role does language difference play for patterns of in uence? The new work order is also characterized by extended networking. Large companies 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 analyse if and how small network partners manage to in uence 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 nd a balance in policy and practice between these various considerations. Important topics for future research would be to analyse how discourse in organizations

138 Britt-Louise Gunnarsson and workplaces re ect 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 of cial documents and more informal discourse within organizations?

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 (Sprket, 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 number 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 presentation of the case studies of second language speakers at a Swedish hospital is found in Andersson (2009), which is her Ph.D. thesis. Nelsons 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).

Andersson, H. (2006), Interkulturell kommunikation inom sjukvrden. Den kommunikativa situationen fr invandrare p svenska arbetsplatser 1. Intervjuer. (TeFa nr 45). Uppsala.

Discourse in Organizations and Workplaces


Andersson H. (2009), Interkulturell kommunikation p ett svenskt sjukhus. Fallstudier av andrasprkstalare i arbetslivet [Intercultural Communication at a Swedish Hospital. Case Studies of Second Language Speakers in a Workplace]. (Skrifter utgivna av Institutionen fr nordiska sprk 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. 2746. Bargiela-Chiappini, F. and Nickerson, C. (eds) (1999), Writing Business: Genres, Media and Discourses. New York: Longman. Brestam, U. (2005), Culture and culture again. An inter-Nordic merger in the banking sector, in B.-L. Gunnarsson (ed.), Communication in the Workplace. (TeFa nr 42), Uppsala, 6077. 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), 466470. Fant, L. (1992), Scandinavians and Spaniards in negotiation, in A. Sjgren and L. Janson (eds), Culture and Management. In the Field of Ethnology and Business Administration. Botkyrka, pp. 125153. 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), 139188. 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. 5169. Gunnarsson, B.-L. (2004a), The multilayered structure of enterprise discourse, Information Design Journal + Document Design, 12, (1), 3648. 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. 1739. 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. 1317. Gunnarsson, B.-L. (2006), Swedish companies and their multilingual practices, in J. C. Palmer-Silveira, M. F. Ruiz-Garrido and I. Fortanet-Gmez (eds), International and Intercultural Business Communication. Theory, Research and Teaching. Bern: Peter Lang, pp. 243263. Gunnarsson, B.-L. (2009a), Professional Discourse. (Continuum Discourse Series.) London: Continuum. Gunnarsson, B.-L. (2009b), Multilingualism within transnational companies. An analysis 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: PalgraveMacmillan. Gunnarsson, B.-L. and Jmtelid, K. (1999), versttning och parallellskrivning ersprkig textproduktion inom ett internationellt storfretag, in Svenskans beskrivning 23. Lund: Lund University Press, pp. 134142.

140 Britt-Louise Gunnarsson

Holmes, J. (2000), Politeness, power and provocation: How humour functions in the workplace, Discourse Studies, 2, (2), 159185. Holmes, J. (2005), Story-telling at work: A complex discursive resource for integrating personal, professional and social identities, Discourse Studies, 7, (6), pp. 671700. 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 Institutional Order. Discourse in Medical, Mediation and Management Settings. Berlin and New York: Mouton de Gruyter, pp. 351385. 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: Identifying virtues and emotions in an emerging business genre, in A. Trosborg and P. E. Flyvholm Jrgensen (eds), Business Discourse. Texts and Contexts, Bern, Berlin, Frankfurt am Main: Peter Lang, pp. 111138. Johansson, C. (2003), Visioner och verkligheter. Kommunikationen om fretagets strategier. Department of Information Sciences, Uppsala University, Uppsala. Jmtelid, K. (2001), Multilingual text production at an international company, in F. Mayer (ed.), Language for Special Purposes: Perspectives for the New Millenium. Volume 2. Tbingen: Gunter Narr Verlag, pp. 797805. Jmtelid, K. (2002), Texter och skrivande i en internationaliserad affrsvrld. Flersprkig textproduktion vid ett svenskt storfretag. 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. 4259. Kankaanranta, A. (2005b), Hej Seppo, could you pls comment on this! Internal Email Communication in Lingua Franca English in a Multinational Company. Centre for Applied Language Studies, University of Jyvskyl, Jyvskyl. Karlsson, A. M. (2006), En arbetsdag i skriftsamhllet. Ett etnogra skt perspektiv p skrifanvndning i vanliga yrken. (Smskrift utgiven av Sprkrdet). Stockholm: Norstedts akademiska frlag. Karlsson, A.-M. (2007), Text, situation, praktik. Om ramar och resurser fr tolkning av texter i arbetsrelaterade skrifthndelser, in B.-L. Gunnarsson and A.-M. Karlsson (eds), Ett vidgad textbegrepp. (TeFa nr 46). Uppsala, 2740. Kong, K. C. C. (2002), Managing the ambiguous and con icting identities of upline and downline in a network marketing rm, Discourse Studies, 4, (1), 4974. Landqvist, H. (2001), Rd och ruelse. Moral och samtalsstrategier i Giftinformations centralens telefonrdgivning. (Skrifter utgivna av Institutionen fr nordiska sprk 55.) Uppsala. Lemke, J. L. (1999), Discourse and organizational dynamics: Website communication and institutional change, Discourse & Society, 10 (1), 2147. LeVine, P. and Scollon, R. (ed.) (2004), Discourse and Technology: Multimodal Discourse Analysis. Washington, DC: Georgetown University Press.

Discourse in Organizations and Workplaces


Linde, C. (1999), The transformation of narrative syntax into institutional memory, Narrative Inquiry, 9, (1), 139174. Louhiala-Salminen, L. (1995), Drop Me a Fax, Will You?: A Study of Written Business Communication. (Reports from the Department of English 10). Jyvskyl: University of Jyvaskyla, 1115. Luzon, M. M. J. (2002), A genre analysis of corporate home pages, LSP and Professional Communication, 2 (1), 4156. Marra, M. and Holmes, J. (2004), Af liation workplace narratives and business reports: Issues of de nition, Text, 24, (1), 5978. Nelson, M. (2007), Professionella och socialiserande kommunikativa aktiviteter. Andrasprkstalande kvinnor och mn i arbete, in B.-L. Gunnarsson, S. Entzenberger and M. Ohlsson (eds), Sprk och kn i nutida och historiskt perspektiv. (Skrifter utgivna av Institutionen fr nordiska sprk 71.) Uppsala, 287298. Nelson, M. and Andersson, H. (2005), Andrasprkstalare i arbetslivet. Data frn intervjuer och fallstudier, in U. Brestam and B.-L. Gunnarsson (eds), Sprk och kultur i det multietniska Sverige (TeFa nr 44), Uppsala, pp. 8497. 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, Communication and the Enterprise. Linguistic Perspectives, University of Lisbon Centre for English Studies, Lissabon, pp. 6578. 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 Americas 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. 4363. 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), 223242. Wiberg, M. (ed.) (2005), The Interaction Society. Practice, Theories and Supportive Technologies. Hershey, London, Melbourne, Singapore: Information Science Publishing.


8 Political Discourse and Translation

Christina Schffner

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 ful l functions for political communication across languages, cultures and ideologies. This chapter addresses the (in)visibility of translation in political communication, textual pro les 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/templates/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 presidents 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 mistranslation. For example, the New York Times published an article that focused exclusively on the translation or possible mistranslation of the statement in order to determine 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 translation 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 re ects an assumption of monolingualism being the normal form of communication within a society. In a positive sense, translation is immediately visible in the eld 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 nancial markets, 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 Schffner 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 common diplomatic practice, with Merkel speaking in German and Sarkozy in French. In another example, the British weekly magazine The Economist in its regular comment 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 Putins 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 re ects a general tendency of accepting translations as authentic 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 assumption is that the translation is an exact replica of the original text. This deep-seated expectation of faithfulness and delity 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 expectation 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. Blairs reaction to Ahmadinejad, quoted above). Media columnists structure their own texts around statements by politicians presented in translation (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) Russias president, Vladimir Putin, said he would support Dmitry Medvedev, rst 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, meaning that Mr Putin is likely to call the shots. (The Economist, 15 December 2007: 9)

Political Discourse and Translation 145 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 ful l 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 de ned. As argued elsewhere (Chilton and Schffner, 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 de ned as potentially political (Chilton and Schffner, 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 (discursive practices). Burkhardt (1996) suggests a broad distinction between communicating 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 institutional politics and everyday politics, this chapter only addresses institutionalized forms of political discourse, i.e. texts that originate in political or media institutions. 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 signi cant 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 political aims and decisions to the outside world. But political aims and decisions of foreign 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

146 Christina Schffner United Kingdom is not of cially a multilingual country (cf. Blackledge, 2005), a variety of documents are translated into community languages. For example, a lea et 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 lea et includes a section on helping to prevent a terrorist attack, which means that this lea et 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 contexts, causes and effects of translations can be discovered (cf. Chesterman, 1998). In the following sections, I illustrate the link between translation pro les 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; Schffner, 2005; on previous work on the topic of translation and politics see Schffner, 2007).

8.3 Translation Pro les 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 of cial languages English and French. For the European Union, all national languages of the member states are of cial languages due to the EUs 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: intracultural communication and intercultural communication, depending on the producers, addressees and functions of the texts (e.g. intrainstitutional translation

Political Discourse and Translation 147 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 signi cant number of of cial texts, the principle of equal authenticity of all languages and texts applies. This policy has led to the phenomenon that translation, although a huge enterprise, is invisible in the of cial discourse (e.g. the Council Regulation No 1 speaks of language versions, but not explicitly of translation). The political aim of equal authenticity results in speci c 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 language 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 Prambel Die Vlker 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-religisen und sittlichen Erbes grndet sich die Union auf die unteilbaren und universellen Werte der Wrde des

148 Christina Schffner Menschen, der Freiheit, der Gleichheit und der Solidaritt. Sie beruht auf den Grundstzen der Demokratie und der Rechtsstaatlichkeit. Sie stellt den Menschen in den Mittelpunkt ihres Handelns, indem sie die Unionsbrgerschaft und einen Raum der Freiheit, der Sicherheit und des Rechts begrndet. (http://eur-lex.europa.eu/JOHtml.do?uri=OJ:C:2004:310: SOM:DE:HTML) (3c) Partie II La Charte des Droits Fundamentaux de lUnion Prambule Les peuples dEurope, en tablissant entre eux une union sans cesse plus troite, ont dcid de partager un avenir paci que fond sur des valeurs communes. Consciente de son patrimoine spirituel et moral, lUnion se fonde sur les valeurs indivisibles et universelles de dignit humaine, de libert, dgalit et de solidarit; elle repose sur le principe de la dmocratie et le principe de ltat de droit. Elle place la personne au cur de son action en instituant la citoyennet de lUnion et en crant un espace de libert, de scurit et de justice. (http://eur-lex.europa.eu/JOHtml.do?uri=OJ:C:2004:310: SOM:FR:HTML) (3d) Parte II Carta de los Derechos Fundamentales de la Unin Prembulo Los pueblos de Europa, al crear entre s una unin cada vez ms estrecha, han decidido compartir un porvenir pac co basado en valores comunes. Consciente de su patrimonio espiritual y moral, la Unin est fundada sobre los valores indivisibles y universales de la dignidad humana, la libertad, la igualdad y la solidaridad, y se basa en los principios de la democracia y el Estado de Derecho. Al instituir la ciudadana de la Unin y crear un espacio de libertad, seguridad y justicia, sita a la persona en el centro de su actuacin. (http://eur-lex.europa.eu/JOHtml.do?uri=OJ:C:2004:310: SOM:ES:HTML) As said above, it is the language policy of the EU which requires a close similarity of the different versions. The dif culty, or rather, the impossibility, to identify which language version might have been based on which other one can be illustrated with reference to the use of the heart metaphor in the English and French versions (at the heart of its activities, au cur), 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 speci c with geistig-religisen und sittlichen Erbes (literally: spiritual-religious and moral-ethical heritage) compared

Political Discourse and Translation 149 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 conveying 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 Europes history or not. The decision to add religis 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 member 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 ( %2B+prodi&hl=en&ie=UTF-8; http://www.ambafrance-uk.org/Risks-of-de-industrialization.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-speci c conventions can be seen in the forms of address (Sehr geehrter Herr Prsident, Dear President, Monsieur le Prsident) and the closing formula (Mit freundlichen Gren, Yours sincerely, Nous vous prions dagrer, Monsieur le Prsident, lexpression de notre haute considration). In the introduction to the letter, the sequence in listing the names is different (for the German version: Jacques Chirac, Gerhard Schrder, Tony Blair; for the English and French versions: Jacques Chirac, Tony Blair, Gerhard Schrder, 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 rst (the French version did not repeat the names at the end). (4a) In einem gemeinsamen Brief wandten sich Staatsprsident Jacques Chirac, Bundeskanzler Gerhard Schrder und Premierminister Tony Blair bei ihrem Treffen am 20. September 2003 in Berlin an den Prsidenten der Europischen Kommission, Romano Prodi: Sehr geehrter Herr Prsident, der Europische Rat hat auf unsere Initiative am 21. Mrz 2003 ein klares Signal zur Strkung der industriellen Wettbewerbsfhigkeit 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 mssen, entscheidend verbessern mssen. [. . .] Mit freundlichen Gren Gerhard Schrder 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 Schrder, Chancellor of

150 Christina Schffner 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 Schrder (4c) Risques de Desindustrialisation Lettre du Prsident de la Rpublique, M. Jacques Chirac, du Premier Ministre du Royaume-Uni, M. Tony Blair, du Chancelier de la Rpublique Fdrale dAllemagne, M. Gerhard Schroeder, au Prsident de la Commisson Europenne, M. Romano Prodi (Berlin, 20 septembre 2003) Monsieur le Prsident, A notre initiative, le Conseil europen a, le 21 mars 2003, mis un signal clair en faveur du renforcement de la comptitivit industrielle de lUnion. Nous avons reconnu la ncessit de rduire la charge bureaucratique qui pse sur les entreprises europennes et damliorer de manire dcisive le cadre rglementaire dans lequel celles-ci, confrontes une vive concurrence, doivent oprer. [. . .] Nous vous prions dagrer, Monsieur le Prsident, lexpression de notre haute considration. 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 relatively close similarities between the English and French versions (e.g. sent out a clear signal and mis un signal clair; both starting in exactly the same way with the prepositional phrase on our initiative/ notre initiative, European companies/les entreprises europennes 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 ful l a speci c and shortterm 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.

Political Discourse and Translation 151

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 original 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 embassies often publish more speeches, both in the original and in translation, than governments 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 rati ed participants in an exchange and who are directly addressed), auditors (those who are known, rati ed, but not directly addressed), overhearers (known but not rati ed participants and not addressed), and eavesdroppers (those whose presence is not even known). The two examples below are a New Years message and a statement in Parliament. The rst case is Tony Blairs New Years 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 identi ed 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 ful lling 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 speci c to the UK culture were taken into account for producing the translation, as can be seen in translation strategies of explanation, explicitation, information addition. For example, in a section on economic growth, Bank of England, windfall tax and New Deal have been explained and/or speci ed 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 scal rules, two years of tight spending, the windfall tax on the privatized utilities paying for the New Deal.

152 Christina Schffner All das verdanken wir nicht dem Zufall, sondern harten Entscheidungen: Unabhngigkeit unserer Zentralbank der Bank von England (literally: our Central Bank the Bank of England ), harte steuerpolitische Manahmen, zwei Jahre strenger Ausgabendisziplin, die windfall Steuer, mit der unerwartete Gewinne aus privatisierten Unternehmen der ffentlichen Versorgung abgeschpft und fr die Politik des New Deal eingesetzt wurden (literally: the windfall tax with which unexpected pro t 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 nd on government websites and embassy websites are normally translated with attention to a uent 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 dif cult of all. In den letzten Jahren haben wir einige sehr harte Entscheidungen getroffen, besonders in den letzten zwlf Monaten. Sie waren notwendig, und ich glaube, sie zeigen erste Erfolge. Die schwierigste von allen war die Entscheidung, Krieg gegen den Irak zu fhren. (Literally: Within the last years we have taken some dif cult decisions, particularly in the last twelve months. They were necessary, and I think they show the rst successes. The most dif cult 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 interesting 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 mchte zunchst Ihnen allen

Political Discourse and Translation 153 fr Ihr Kommen und fr die Einladung danken, heute hier zu sprechen. (Ladies and gentlemen was added) (http://www.britischebotschaft.de/de/news/ items/070523a.htm) The production of uent target texts seems to be the normal institutional practice as well when it comes to translating speeches by politicians, which were delivered 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 uency of the target texts. This can be illustrated with the rst 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 Prsident, Sie gehren zu den Parlamentariern der ersten Stunde im Europischen 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 Europischen Parlament entwickelt mit selbstbewussten Abgeordneten, klaren Parteistrukturen und Fraktionsbildungen. Damit wurde es ein anspruchsvoller, aber nicht mehr wegzudenkender Partner im europischen 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 rst 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 in uence the remarkable rise of a Parliament from its infancy. It has evolved into a very emancipated European Parliament with con 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 Prsident, vous faites partie des parlementaires de la premire heure au Parlement europen. Comme vous venez de le dire, vous avez acquis de lexprience depuis 1979. Il me semble pouvoir dire que vous avez vcu et contribu forger la remarquable ascension dun parlement depuis ses premiers balbutiements. Ce parlement est devenu un Parlement

154 Christina Schffner europen trs mancip, avec des dputs srs deux, des structures de parti et des groupes parlementaires bien d nis. Cest aujourdhui un partenaire exigeant, mais qui ne pourrait plus tre absent du dbat europen. (http://www.bundesregierung.de/nn_5846/Content/FR/Reden/ 2007/02/2007-02-13-rede-merkle-ep-fr.html) These ndings are in line with those of Calzada Prez (2001) who analysed translated speeches in the European Parliament (Spanish English). Her analysis, carried out through surface description, illocutionary explanation and (sociopolitical) perlocutionary explanation, revealed a broad variety of translational shifts which were intended to help target texts to be more readable, thus contradicting Koskinens (2000) ndings 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 rst 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 available 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 regimes brutal crackdown against its own people. Monks, nuns and ordinary citizens took to the streets peacefully in protest at the deterioration of the economic situation 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 whereabouts 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. Mnche, Nonnen und normale Brgerinnen und Brger waren auf die Strae gegangen, um friedlich gegen die Verschlechterung ihrer wirtschaftlichen Lage zu protestieren. Das Regime reagierte mit Gewehren und Schlagstcken. Die genaue Zahl der Todesopfer kennen wir nicht, aber sie liegt wahrscheinlich viel hher, als das Regime zuzugeben bereit ist. Der Aufenthaltsort vieler Verhafteten und deren weiteres Schicksal sind nach wie vor unbekannt. Inzwischen setzt das Regime seine Repressalien fort: Die

Political Discourse and Translation 155 Sicherheitskrfte fhren 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 Schlagstcken 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 brutalit avec laquelle le rgime birman a rprim son propre peuple nous a tous horri s. Des moines, des nonnes et des citoyens ordinaires, descendus paci quement dans la rue pour protester contre la situation conomique du pays, ont t accueillis coups de fusils et de gourdins. Le chiffre exact des morts et des blesss reste inconnu et il est probablement beaucoup plus lev que celui avanc par les autorits. Ltat de sant des nombreux prisonniers suscite linquitude. Et, pendant ce temps, les perscutions continuent: les forces de scurit procdent chaque nuit de nouveaux raids et de nouvelles arrestations. (http://www.ambafrance-uk.org/Renforcer-la-pression-sur-la.html) Since the complete text does not re ect a signi cantly 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 speci c ideological agenda is feminism, with translators as interventionists, 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 deliberate and conscious act of selection, assemblage, structuration and fabrication and even, in some cases of falsi cation, refusal of information, counterfeiting, and the creation of secret codes. (Tymoczko and Gentzler, 2002: xxi) On the basis of the nal example, press conferences, I will show how institutional practices result in a particular way of presenting (and constructing) politicians.

156 Christina Schffner

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 interpreting 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 rst 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, rst of all let me give my thanks to Chancellor Schroeder for hosting this meeting and say what a constructive meeting I found it. [. . .] (http://www.number10.gov.uk/output/Page4508.asp) Even when there is an explicit reference to the use of another language, as in the rst 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:

Political Discourse and Translation 157 (11) Prime Minister: Ladies and Gentlemen I am delighted that my rst 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 dont 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. [. . .] (http://www.number10.gov.uk/output/Page12381.asp) 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 transcripts of the statements in a language other than German are based on simultaneous (or occasionally consecutive) interpreting. As the analysis of a number of press conferences has shown, some revision process is in place for the German governments website. The transcripts re ect that the orally delivered contributions 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 re ects the typical features of oral speech (e.g. false starts, self corrections), whereas in German, Bushs discourse is uent 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 veri able fashion, were 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 begeben? Die Bundeskanzlerin hat mich auch dazu aufgefordert, darber nachzudenken. Ich habe dann Folgendes gesagt: Wenn die Iraner nachweislich mit der Urananreicherung aufhren, dann werden wir zum Verhandlungstisch zurckkehren. (Literally: We do the same in respect

158 Christina Schffner 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 negotiate? The Chancellor asked me to re ect about this. I then said the following: 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 inevitably arrive at different evaluations of the politicians performance. In November 2007, the German Chancellor Merkel visited US President Bush. The press conference 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 rst 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 Prsident and BKin for Bundeskanzlerin, the female form of Federal Chancellor): (13) P Bush: (auf Englisch) BKin Merkel: Sehr geehrter Herr Prsident, lieber George, ich mchte mich zuerst fr diese Mglichkeit 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 conference too must leave the German reader wondering what was going on. The German website provides the following information: P Bush: (auf Englisch) BKin Merkel: Das ist fr mich als eine Hamburgerin natrlich 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 Merkels 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: Im now going to go feed the Chancellor a hamburger. (Laughter.) Right here, Crawford, Texas. No, well, I mean back over there. Thank you all.

Political Discourse and Translation 159 Chancellor Merkel: Obviously, for me, as a person who originally came from Hamburg President Bush: Yes. Chancellor Merkel: its even more important. President Bush: Hamburger. Thank you. Chancellor Merkel: Thank you. Comparing the English version of Merkels 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: BKin Merkel: Sehr geehrter Herr Prsident, lieber George, ich mchte mich zuerst fr diese Mglichkeit bedanken, hier in Texas diese Gesprche zu fhren, auch im Namen meines Mannes. Es ist ein wunderschnes Fleckchen Erde, wie man in Deutschland sagen wrde, das etwas von der amerikanischen Weite und der Vielfalt der Landschaft zeigt. Herzlichen Dank dafr, dass wir heute Morgen die Mglichkeit zum Spaziergang hatten. Das war eine wirklich neuartige Erfahrung. (Literally: Dear Mr President, dear George, rst of all, allow me to thank you for the possibility to have these talks here in Texas, also on behalf of my husband. 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 accompanies 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 Merkels 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

160 Christina Schffner (e.g. a very beautiful spot, a very beautiful part of this planet, of this world) are thus characteristic features of the interpreters 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 Im curious about that, too. (Laughter) President Bush: The pig? Ill 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 darber, [. . .] P Bush: Ich dachte, es wrde 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 speci city 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-speci c competence (cf. Neubert, 2000). Translators operate within speci c 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 ful l for its addressees in the target culture setting. This involves critical re ection about

Political Discourse and Translation 161 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 re ects 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 signi cance 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 comparative 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 salutation formula in letters). For explaining such translation strategies, additional concepts 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 (Schffner, 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, substitution, 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 speci c discursive practice is being recontextualized. There is one nal 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

162 Christina Schffner 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 speci c 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 translations 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 discourse and can thus yield insights into the intricacies of cross-cultural political communication and interaction.

Baker, M. (2006), Translation and Con ict: A Narrative Account. New York and London: Routledge. Bassnett, S. (2004), Trusting reporters: What exactly did Saddam say?. The Linguist, 43, (6), 176178. 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, 145204. Bielsa, E. (2007), Translation in global news agencies, Target, 19, (1), 135155. 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, ffentlichkeit, vol. 7). Berlin and New York: de Gruyter, pp. 75100. Calzada Prez, Ma. (2001), A three-level methodology for descriptive-explanatory Translation Studies, Target, 13, (2), 203239. Chesterman, A. (1997), Memes of Translation. Amsterdam and Philadelphia: John Benjamins.

Political Discourse and Translation 163

Chesterman, A. (1998), Causes, translations, effects, Target, 10, (2), 201230. Chilton, P. and Schffner, 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. 206230. 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. 258284. Holland, R. (2006), Language(s) in the global news: Translation, audience design and discourse (mis)representation, Target, 18, (2), 229259. Kang, J.-H. (2007), Recontextualization of news discourse: A case study of translation of news discourse on North Korea, The Translator, 13, (2), 219242. Koskinen, K. (2000), Institutional illusions: Translating in the EU Commission, The Translator, 6, (1), 4965. Koskinen, K. (2001), How to research EU translation?, Perspectives, 9, (4), 293300. Lilleker, D. (2006), Key Concepts in Political Communication. London: Sage. Mason, I. (2000), Audience design in translating, The Translator, 6, (1), 122. Neubert, A. (2000), Competence in language, in languages, and in translation, in C. Schffner and B. Adab (eds), Developing Translation Competence. Amsterdam and Philadelphia: John Benjamins, pp. 318. Schffner, C. (2004), Political discourse analysis from the point of view of translation studies, Journal of Language and Politics, 3, (1), 117150. Schffner, C. (2005), Bringing a German voice to English-speaking readers: Spiegel International, Language and Intercultural Communication, 5, (2), 154167. Schffner, C. (2007), Politics and translation, in P. Kuhiwczak and K. Littau (eds), A Companion to Translation Studies. Clevedon: Multilingual Matters, pp. 134147. Seymour, E. (2002), A common EU legal language?, Perspectives 10, (1), 714. Simon, S. (1996), Gender in Translation: Cultural Identity and the Politics of Transmission. London: Routledge. Snell-Hornby, M., Pchhacker, 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 Multilingual 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 Martnez, J. M. (2002), Translating for the European Union Institutions. Manchester: St Jerome.


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 (re ection), 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 signi cant concern, because the law is such an important and in uential 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 signi cance for applied linguists is that the linguistic aspects of the law raise many issues and dif culties. The written language of legislation and regulation is dif cult to understand 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 dif culty (the re ection 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 rst 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 ef ciently only by using legal jargon. The fourth is the interpersonal arena, where power disparities and hyperformality are produced by the essentially controlling nature of the legal system. Looking rst at genres, the highly institutionalized, and sometimes ritualized discourse of the law often follows regular patterns: organized sequences of elements 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 understanding (Wallace, 1990; Weaver, 1988). This is in part why legal documents can be dif cult for lay readers to understand, while lawyers have less dif culty. 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: 922931), 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 nds this incompetent and unintelligible. The lawyers, however, 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 concerning 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 witnesss life, loves and previous contacts with the law. This narrative interpretation both facilitates analysis, and problematizes the process. For example, 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 evidence 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 convincing 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 AngloWelsh 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 of cers 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 consequences are far-reaching. They include the possibility of extremely long and complex sentence structures (often between 70 and 100 words), which are virtually 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 metaphor) 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 investigation 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 negotiation of meaning. The problem is that there is psycholinguistic evidence (Felker et al., 1981) that complex sentence and phrasal structures, and the use of grammatical metaphor, make texts dif cult to understand.

Language and the Law 167 Another part of the dif culty of legal language is its technicality. Maley (1994) among others has pointed out that the law consists to some degree of legal concepts, 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 speci c behaviours and entities/people in speci c circumstances, and legal language attempts to spell out precisely what these are, in order to avoid hostile 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 nonlawyers, it clearly has the potential to impair their understanding of and their participation in the legal process. This accentuates the problems of complexity discussed above. One source of this dif culty is that legal discourse may be addressing two audiences, 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 authority of the person using it. Resistance from lawyers and police of cers 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 comprehensible 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 language movements in the United States, Britain and Australia, and in a range of other countries. In Britain they have recently achieved a signi cant success in persuading 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 substitute 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 removalists contract that she discusses some of the legal content is lost in a plain language version. Gibbons (2001) demonstrates some of the problems caused by the linguistic complexity of police cautions, discusses the sometimes 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 speci c checking interactions, and individually crafted responses on the basis of these checks which target individual comprehension and participation 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 interpersonal power is manifested and exerted to a signi cant 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 of cers 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 sergeants courtroom testimony. Police Of cer: I was unable to maintain the light being illuminated. Counsel: To keep the torch on? Police Of cer: 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 speci c right to silence. While judges have the right to speak whenever they wish, lawyers in general must take turns (Atkinson and Drew, 1979). Turn taking in courtrooms has therefore become regulated and institutionalized along power hierarchy lines. Equally, police of cers 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),

Language and the Law 169 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: . . . Id suggest the reason to you, because you dont want everyone to know the little criminal that you are, do you? Thats the reason, isnt it? Isnt it? Your silence probably answers it, but Ill have an answer from you. Thats the reason, isnt it? The core information you are a little criminal is deeply embedded in the grammatical structure, so that it is very dif cult 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 isnt 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 judges summing up can in uence jurors. Critical discourse analysis is an emerging focus. For instance Vasilachis de Gialdino (1997) examined an Argentinian labour reform bill rooted in neoliberalism, describing the language used within labour courts in Argentina, discussion 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: 270271) 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 exibility. There is also a growing debate concerning 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 dif cult to nd thoroughgoing solutions. Some actions taken so far are to alert and educate lawyers and judges to the risks involved in such questioning, 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. OBarr (1982) and OBarr 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 nancial 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 countries 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 legislation is framed. This places considerable demands upon the teachers and curriculum 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 language 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 rst to legal interpreting and translation, we have already touched on the possible disadvantage suffered by minorities who cannot cope easily with the complexity and power of the language of the law. People who have only a limited command 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 quali ed language testers, there is chronic under provision of interpreters 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 available for all second language speakers unless there is evidence from a quali ed

Language and the Law 171 language tester that the person has suf cient command of the courtroom language 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 pro cient 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 development of alternatives such as minority language courtrooms. Berk-Seligson (1990), Hale and Gibbons (1999) and many others have documented the extreme dif culty of providing accurate interpreting in courtroom contexts, where even minor inaccuracies may lower the standards of justice. The conditions that make this process particularly dif cult 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. seor in Spanish), passivization, 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 crossexamination). In many jurisdictions these problems can be exacerbated by inadequate training and sub-professional rates of pay. Clearly adequate resourcing is a basic rst 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 complexity and technicality discussed earlier. For example Vlachopoulos (2004) discusses 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 Greeces 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 discussion 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 con icts around the world (for instance Macedonian in Kosovo, or Kurdish) is the right to use a language for public 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 legislation to prevent the use of minority languages for public purposes (Gonzlez 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 suf ciently into account the evidence that for children a high level of bilingualism is viable and achievable, while for adult migrant learners of the second language high levels of pro ciency 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 dif culty 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 bribers 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 mans contributions to the discussion, including markers of cooperation (such as yeah), answers to questions, topic management, clari cation requests, interruptions, turn taking, as well as incomprehension markers, that he never actively participated in the deal

Language and the Law 173 itself. The evaluation of the success of such intervention by forensic linguists is usually whether their evidence is accepted by the court, and is af rmed 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 suf cient defence to a prosecution for an offence under this section if the defendant satis es 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 ne 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 participants 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 someone drops a hammer on their foot. On re ection, the extreme discretion within this legislation is dangerous, since in 1993 Amnesty International reported that it was used overwhelmingly to imprison or ne 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 welldocumented instances of police of cers and politicians using swearwords themselves. Indeed one aboriginal man was prosecuted for saying to police Dont tell me to get fucked. Even if one deplores the use of offensive language, in these circumstances 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 vili cation. This can take the form of libel or slander of individuals, or of group vili cation. 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 de nitions of slander in normal dictionaries normally include intent, for instance the Oxford Dictionary de nition 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 considerations 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 vili cation is mostly legislated against in terms of ethnicity, but there is also legislation against vili cation on the grounds of religion, disability or sexual orientation. Group vili cation 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 vili cation, 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 vili cation in place of overt expression.

9.6 Linguistic Evidence

Finally, another major area where language intersects with the law is that of linguistic 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 eld, and if accepted, his/her evidence will be taken without further demur in subsequent court cases. In Common Law systems the competence 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 scienti cally derived. Bowe and Storey (1995: 188189) 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 identi cation. We will examine these in turn, beginning with communication, and moving through the linguistic levels.

Language and the Law 175

9.6.1 Communication
Looking rst 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 dif cult to crack the oral code used in expressions such as [bpkpos pin pu kpappu: pov mpuns]. It emerged that every vowel has an [p] 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 dif cult to understand. Levi (1994: 1617) discusses 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 jargon, crucial terms left vague or unde ned, needlessly complex syntax, anaphora (eg. demonstrative pronouns like this) with obscure antecedents, related information 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 Levis 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 testi ed on behalf of parents whose child died of Sudden Infant Death Syndrome (SIDS) at the age of 18 months. The childs life was protected by the fathers life and accident insurance, which included the statement The plan pays a bene t 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. McMenimans reading of the medical literature and dictionary de nitions revealed that syndrome is distinguished 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 speci c time in a speci c 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 person 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 classi ed as a disease. McMenimans 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 interpretation) 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 Solans 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 (womens/mens business). The consequence 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 investigation and court trials are largely dependent on the right to question and the obligation to answer. Eades has testi ed in court on a number of occasions concerning the resulting lack of communication.

9.6.2 Identi cation

Identi cation may involve comparing two or more language samples, and saying whether they were produced by the same person or not; alternatively it may involve pro ling 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 identi cation 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 pro les, and a single speaker produced markedly different pro les. 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 identi cation, provided that the recording on

Language and the Law 177 which they are based is of adequate quality. Hollien notes, however, that there are many points where recording quality can be lost. Identi cation becomes particularly dif cult 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 background 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 indicator than a machine, particularly when, for example, one is attempting to distinguish 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 effective. Courts still tend to prefer machine based analysis, since it is more overtly scienti c. The untrained ear is unreliable in voice identi cation, earwitness identi cation being even less reliable than eyewitness identi cation. 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 differences. The evidence was accepted and Prinzivalli was acquitted. Similar in nature are various identi catory elements of the written language. Handwriting can be distinctive, and peculiarities of spelling and punctuation can be strong identi ers. In a recent case I was able to pro le a writer as probably coming 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 processes 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 established 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 identi cation or pro ling. The rst 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 structure 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 pro ciency 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 discussed earlier. This phenomenon can be detected and revealed by a linguist. I have testi ed 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 womans husband, who was suspected of murdering her. Eagleson compared samples of the husbands and the wifes writing, and was able to show a range of features that were found in the mans writing and the disputed letter, but not in the womans 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 punctuation 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 coming 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 grammatical patterns. There were also indications from the patterning and format of the letter that the writer may have had some experience in writing radio advertisements (there were signs of intertextuality). This type of pro ling, while it is not conclusive in its identi cation, 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 discourse 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 rst name, or their rst 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 identi ed on linguistic grounds later confesses to producing the

Language and the Law 179 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 important and fast developing area of real-world applied linguistic concern. All the issues discussed here are of major signi cance 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 vili cation 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, scienti c 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 unnecessarily coercive courtroom questioning, particularly where vulnerable witnesses are involved (see Eades, 2000 and 2008).

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. 136155. 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 identi cation: Demand and response, in D. Eades (ed.), Language in Evidence Issues Confronting Aboriginal and Multicultural Australia. Sydney, NSW: University of New South Wales Press, pp. 187200.

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. 231247. Carroll, J. (1995), The Use of Interpreters in Court. Forensic Linguistics, 2, (1), 6573. 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. 414427. Coulthard, M. (1997), A failed appeal, Forensic Linguistics, 4, (2), 287302. 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. 100135. 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. 222234. 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. 8299. Diamond, S. S. and Levi, J. N. (1996), Improving decisions on death by revising and testing jury instructions, Judicature, 79, (5), 224232. Eades, D. (1994), A case of communicative clash: Aboriginal English and the legal system, in J. Gibbons (ed.), Language and the Law. London: Longman, pp. 234264. 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. 181190. Eades, D. (2000), I dont think its an answer to the question: Silencing Aboriginal witnesses in court, Language in Society, 29, 161195. 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. 179195. Eagleson, R. (1994), Forensic analysis of personal written texts: A case study, in J. Gibbons (ed.), Language and the Law. Harlow: Longman, pp. 363373. 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), 439469. Gonzlez, R. D. and Melis, I. (eds) (2001), Critical Perspectives on the Of cial 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. 215229.

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. 83111. 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 Courtroom. Oxford: Blackwell. Solan, L. M. (1995), Judicial decisions and linguistic analysis: Is there a linguist in the court? Washington University Law Journal, 73, (3), 10691083. 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. 219258. 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 Poltico y Prensa Escrita. Barcelona: Editorial Gedisa. Vlachopoulos, S. (2004), Translating the untranslatable? The impact of cultural constraints 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. 100115. 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. 203218. Walsh, M. (1994), Interactional styles in the courtroom, in J. Gibbons (ed.), Language and the Law. Harlow: Longman, pp. 217233. 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.

Language and the Law 183

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 Multicultural Australia. Sydney: University of New South Wales Press.

Vili cation
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 Identi cation. New York: Academic Press. Coulthard, M. and Johnson, A. (2008), An Introduction to Forensic Linguistics Language in Evidence. London: Routledge.


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 eld try to understand the way language is organized in the brain, taking the performance of people with neurological impairments as the main source of evidence (Whitaker and Whitaker, 19761979; Goodglass, 1993; Caplan, 1987; Ahlsen, 2006). The conceptual basis of this eld 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 suffered strokes in their 60s. In addition, they had long-stay chronic care arrangement in medical centres with interdisciplinary teams of health care and research professionals. This particular medical context created the opportunity to make detailed observations of spared and impaired language, speech and voice functions in people over a long period of time and correlate this with detailed neurological and psychological assessments. For the next 30 years, groups of stroke patients were studied by clinical researchers who had the bene ts of insights from Chomskys 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 con uence of opportunities. Bedside examinations, which were the rule-of-thumb approach used by clinical neurologists, have given way to formal assessment techniques developed by psychologists 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 developing in tandem with other clinical assessments. From the advent of radioactive scanning (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 con rmation 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 pro le observed in a narrowly de ned 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 re ects 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 identi ed as crucial to our understanding 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 re nements in both the way research questions are framed and the manner in which answers are sought. Any domain of active research investigation undergoes re nement over the years, from observation to classi cation, de ning categorical distinctions and generating hypotheses. With the move to experimental research, re nement is achieved through the identi cation of relevant sources of variation in populations and task properties which affect the nature of observations. Understanding 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 characterize 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 de ned as age, education level, occupation, handedness, etiology

186 Marjorie Lorch (i.e. cause of illness) and time post onset of illness, with hearing disorders, developmental speech dif culties and psychiatric problems excluded. The neuropathological evidence was supplied by electroencephalographic recordings (EEG). The interpretative meaning of the rst subject variable that of age has grown in signi cance 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 knowledge 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 significant issue with regard to behaviour and the brain. Towards 1900, there was increasing 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 re nements 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 different 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 adequate account. New understanding of genetics on the one hand, and increasing research on the healthy aging brain with the recent extension in lifespan expectancy 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 neurolinguistics was based on adults with acquired disorders, new research on Speci c Language Impairment (Gopnik and Crago, 1991) and other developmental disorders which affect language such as Williams Syndrome (Bellugi et al., 1997), Down Syndrome and Autism (Tager-Flusberg, 1994) provide another source of important evidence (Jenkins, 2000). Although some of this research is carried out with groups of individuals de ned 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 demonstrated 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 ndings imply a level of linguistic development far beyond that which would be predicted by the childs mental age of one and a half years. The observations in such a case indicate the possibility of an atypical route to language

Neurolinguistics and the Non-monolingual Brain 187 acquisition and the development of literacy for which existing cognitive accounts are inadequate. In order to account for this childs 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 identi cation 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 suf ciently grounded in neurolinguistics to identify atypically behavioural phenotypes or symptom complexes.

10.3 The Variable of Time Post-Onset of Illness

The focus in traditional neurolinguistic research has been on adult acquired neurogenic disorders of language. The majority of aphasic subjects studied had suffered cerebrovascular accident after acquiring language normally. In the twentieth century, this was a major source of illness in Western populations. (In Schuells case series the frequency was 83 per cent, Schuell et al., 1964.) Properties of this neurovascular insult de ne 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 rst two months and chronically. The understanding of how recovery unfolds over time in response to a cardiovascular 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 organization of language through these distinct patterns of impairment. Alzheimers disease (Nicholas et al., 1985), Primary progressive dementia (Hodges et al., 1992), Multiple Sclerosis (Devere et al., 2000), Parkinsons disease (McNamara and Durso, 2003), Spasmodic Dysphonia (Whurr et al., 1993), etc., all provide sources of evidence 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 signi 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 neurolinguistic 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 neurolinguistic 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 impairments in communication in people suffering right hemisphere damage (Code, 1987). This line of research has been assisted by the growing elds 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 in uences.

10.4 The Language Variable

Although the research landscape reviewed above showed a positive trend in potential 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

Neurolinguistics and the Non-monolingual Brain 189 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 ( rst 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 description 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 emerging 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 century has contributed to more published research on language disorders in people speaking languages other than English. In parallel, there has been a owering 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. Throughout 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, different 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 dif culties and posterior lesions with comprehension problems), the patterns of impairment documented in non-monolinguals are dif cult 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 languages 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 patients native language 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 rst because it is the one that uses the most solidly xed 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 relevance to the picture of language impairment in non-monolingual speakers: (1) the language of the patients present environment, (2) the individuals communicative needs, (3) the individuals 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 difference between the grammars of the languages spoken etc. Paradis (1977) devised a classi cation of bilingual aphasia recovery patterns: Parallel similarly impaired and recover at the same rate Differential impairment and recovery of a different degree in the different languages relative to the pre-morbid mastery Successive one language does not begin to reappear before another is maximally recovered Antagonistic one language regresses as another progresses Selective recovery occurs in one language but not in another Blended or mixed inappropriate mixing of two or more languages (not equivalent to normal code switching) 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

Neurolinguistics and the Non-monolingual Brain 191 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 identi ed could account for the observations: Neither primacy, automaticity, habit strength, stimulation pre- or post-onset, appropriateness, need, affectivity, severity of aphasia, type of bilingualism, 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 computational 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 second 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 neuroanatomically 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 dif culty 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 identi ed. 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 pro ciency and attainment and aspects of literacy are also relevant factors. All of these factors shape individual language histories which will be re ected in different neurolinguistic instantiations. In addition, differential use of languages on a day-to-day basis may also lead to signi cant 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 patients le, 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 eld of bilingual aphasia research have been discouraging. In a review of the current state of neurolinguistics, Ahlsen states Neurolinguistic aspects of bi or multilingualism have only been studied extensively by a few researchers (2007: 127). In another recent overview of research ndings in this area, Fabbro (2002: 202203) asserted that, It can thus be concluded that so far empirical studies have not provided tenable explanations for the presence of parallel 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 forward, 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 localization of languages in the healthy bilingual brain. For example, Mechelli and colleagues (2004) claim that the grey matter in Brocas 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 changing 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 imaging 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 bilinguals differing in age of onset of language exposure, thereby making this source of evidence not very informative about individual differences in brain organization related to language experience. (Hull, 2003: 12) This underscores again the importance of determining the theoretical status of such subject variables with precision.

Neurolinguistics and the Non-monolingual Brain 193

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 compatible with understanding that is gained through (1) developments in anatomy and physiology to characterize the function of the nervous system, (2) social and cognitive 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 undocumented because investigators had not thought to examine them. Stimulus materials, task design and assessment materials must re ect 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 properties which means that they may not represent the functioning of language systems 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 representation, given that what makes language most speci c 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 dif culty of learning new words but not procedures by anterograde amnesic patients such as H.M., (4) the dif culties of word retrieval but not syntax in those suffering from Alzheimers dementia, (5) loss of lexical but not syntactic performance in rst language attrition, (6) the relative ease of vocabulary learning in latesecond-language students relative to dif culty 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 explanations regarding language processes per se. Consider the currently widely used clinical task of word uency. 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 diagnostic tool with patients suffering from a diverse group of disorders which involve cognitive but not language impairments (e.g. Neuro bromatosis 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 theoretically coherent for an individual pattern of spared and impaired behaviours in acquired disorders. By generating subject variables which are theoretically derived, more convergence of ndings 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 nd solutions. For example, bilingual aphasia appears to be a crucial area of neurolinguistics in which to address questions about biobehavioural concomitants of language experience. However, as the review presented above demonstrates, current researchers are at a loss to provide a coherent explanation of the body of research ndings 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 Tourettes syndrome is a ground-breaking piece of research. The development of vocal tics and involuntary swearing does not currently have a theoretically motivated 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 identi ed post-infectious streptococcus antibodies as one source of the neuropathology leading to Tourettes 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

Neurolinguistics and the Non-monolingual Brain 195 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 bilingual aphasia using an applied history approach. The aim is to reveal some new avenues for research into language disorders in non-monolingual speakers. Paradis (1983) extensive review of the historical literature identi ed a paper by Pitres (1895) as being the rst 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 identi ed 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 differently by nineteenth-century practitioners. These cases were not clinically classied 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 languages was considered to have a different psychological status than today. They were assumed to be a re ection 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 rst language acquisition of a native language and other types of learning. In order to develop our ideas about language organization in non-monolinguals we may need to revise our assumptions about the contribution of learning and memory. With our growing understanding 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 language 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 anthropological perspective, is more likely to be multilingual and illiterate. Nevertheless, a large portion of modern research into language processing is focused on the neurolinguistic 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 Barrire, 2002, 2003). In the class of acquired disorders of cognitive functions, pure agraphia represents something of an anomaly. It was rst 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 awed piece of outlier data (Barrire 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 signi cant 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 Brocas area at his death. This case, published by Thomas Barlow in 1877, was cited in the literature for over three decades as de nitive 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 impairment and the size and locations of the lesions. These were compounded by subsequent authors when citing this case. This dubious evidence was nevertheless used by leading theoreticians to support arguments about language function, development of dominance and recovery patterns in children. These notions were so pervasive in the late nineteenth and early twentieth century that this one, poorly reported case was suf cient 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 todays clinical researchers. Other issues which have been illuminated by this type of applied historical neurolinguistics are the evolving

Neurolinguistics and the Non-monolingual Brain 197 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 identi ed (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. However, 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 neurologically impaired and healthy adults and children. There has been only limited progress in the interpretation of these ndings 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 eld.

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, 117. Baron-Cohen, S. (2002), The extreme male brain theory of autism, Trends in Cognitive Sciences, 6, 248254. Barrire, I. and Lorch, M. (2004), Premature thoughts on writing disorders, Neurocase, 10, 91108. 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: Cambridge University Press, pp. 328374. 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, 344342. Benson, D. F. (1967), Fluency in aphasia: Correlation with radioactive scan localization, Cortex, 3, 373394. 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 magnetoencephalography, Neurocase, 12, 322331. Caplan, D. (1987), Neurolinguistics and Linguistic Aphasiology: An Introduction. Cambridge: Cambridge University Press. 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), 359372. Chiarello, C., Knight, R. and Mandel, M. (1982), Aphasia in a prelingually deaf woman, Brain, 105, 2951. 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, 12071209. Duffy, F. (1985), The BEAM Method for neurophysiological diagnosis, Annals of the New York Academy of Sciences, 457, (1), 1934. 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, 201222. 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 mechanisms, associations, and pathology: I. A hypothesis and a program for research, Archives of Neurology, 42, 428459. 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, 150. Green, D. (1986), Control, activation, and resource: A framework and model for the control of speech in bilinguals, Brain and Language, 27, 210223. Green, D. (1993), Towards a model of L2 comprehension and production, in R. Schreuder and W. B. (eds), The Bilingual Lexicon. Philadelphia: John Benjamins, 249277. Green, D. (1998), Mental control of the bilingual lexico-semantic system, Bilingualism 1, 6781. Grodzinsky, Y. (1990), Theoretical perspectives on language de cits, 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 Neurology: Considerations of Recovery Patterns in Acquired Childhood Aphasia. In Finger, S. Boller, F. and Tyler, K.L. (eds.) Handbook of Clinical Neurology, Vol 95 (3rd series) History of Neurology. Elsevier, Edinburgh, 2009, 84552.

Neurolinguistics and the Non-monolingual Brain 199

Hillis, A. E., Barker, P., Beauchamp, N., Winters, B., Mirski, M. and Wityk, R. (2001), Restoring blood pressure reperfused Wernickes area and improved language, Neurology, 56, 670672. Hodges, J. R., Patterson, K., Oxbury, S. and Funnell, E. (1992), Semantic dementia: Progressive uent aphasia with temporal lobe atrophy, Brain, 115, (6), 17831806. Hull, R. G. (2003), How does bilingualism matter? A meta-analytic tale of two hemispheres. 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 signi cance of polyglot aphasia, Cortex, 43, 658666. Lorch, M. (2006b), Phrenology and Methodology, or playing tennis with the net down. Aphasiology, 20, 10591071. Lorch, M. (2006c), Language and memory disorder in the case of Jonathan Swift: considerations on retrospective diagnosis, Brain, 129, 31273137. Lorch, M. (2008), Multiple languages, memory and regression: An examination of Ribots law, Aphasiology, doi:10.1080/02687030801931182 Lorch, M. and Barrire, 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. 193205. Lorch, M. and Barrire, I. (2003), The history of written language disorders: reexamining Pitres case (1884) of pure agraphia, Brain and Language, 85, 271279. Lorch, M., Ferner, R., Golding, J. and Whurr, R. (1999), The nature of speech and language impairment in adults with neuro bromatosis, Journal of Neurolinguistics, 12, 157165. McNamara, P. and Durso, R. (2003), Pragmatic communication skills in Parkinsons Disease, Brain & Language, 84, 414423. Mechelli, A., Crinion, J. T., Noppeney, U., ODoherty, 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 fr Neurologie und Psychiatrie, 21, 4372. Naeser, M. and Hayward, R. (1978), Lesion localization in aphasia with cranial computed tomography and the Boston Diagnostic Aphasia Exam, Neurology, 28, 545. Nicholas, M., Obler, L., Albert, M. and Helm-Estabrooks, N. (1985), Empty speech in Alzheimers disease and uent aphasia, Journal of Speech and Hearing Research, 28, 405410. Paradis, M. (1977). Bilingualism and aphasia, in H. Whitaker and H. Whitaker (eds), Studies in Neurolinguistics (Vol. 3, pp. 65121). 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, 178180. Paradis, M. (2001), Bilingual and Polyglot Aphasia. Handbook of Neuropsychology (Second Edition. Volume 3). Oxford: Elsevier Science, pp. 6991. 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]. Spring eld, IL: Charles C. Thomas. Pitres, A. (1895), Etude sur laphasie chez les polyglottes. Revue de mdecine 15, 873899. Translated into English by M. Paradis, in M. Paradis (ed.), Readings in Aphasia in Bilinguals and Polyglots. Montreal: Didier, Vol. 15, 873899. 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), 953956. Ribot, T. (1881), Les maladies de la mmoire. Paris: Alcan. Diseases of Memory, 1882. (English translation of Les maladies de la mmoire, 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 cant do for us, and why, Brain and Language, 21, 38. 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.

Neurolinguistics and the Non-monolingual Brain 201

Vargha-Khadem, F., Gadian, D. G., Copp, A. and Mishkin, M. (2005), FOXP2 and the neuroanatomy of speech and language, Nature Reviews Neuroscience, 6, 131137. Weisenburg, T. and McBride, K. E. (1935), Aphasia. New York: Commonwealth Fund. Whitaker, H. and Whitaker, H. A. V. (eds) (19761979), Studies in Neurolinguistics (Vol. 14). 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, 92111.


11 Clinical Phonology
Martin J. Ball and Nicole Mller

11.1 Introduction
Clinical Linguistics might be thought of as a classic example of the use of linguistics in the real world, where the interaction of linguistic theory and clinical data allows a classi cation of disorder type, an analysis of disorder severity and a pathway 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 production 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, theoretical phonology has not remained static since the major breakthrough of Chomsky and Halles 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 linguistics 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 classic SPE model include developments with features, e.g. feature geometry (Clements, 1985), and underspeci cation (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 phonology (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 neurolinguistic and/or psycholinguistic aspects of speech production. This dichotomy we refer to as the difference between a descriptively adequate and psycholinguistically adequate 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 behaviours 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 preferred 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 adequacy 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 produce more elegant descriptions of phonological phenomena. Nevertheless, even within the Generative school, approaches such as underspeci cation (Archangeli, 1988) might be justi ed by appeals to psycholinguistic plausibility. It was argued, e.g., that underspeci ed segments would take less memory storage space than fully speci ed ones (see Spencers 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 Mller psycholinguistic aspect: childrens substitution patterns can be seen as coming about from their underspecifying underlyingly two or more sounds that are contrastive in the target system with that feature being lled in by default (Dinnsen and Barlow, 1998: 71). Other phonological approaches have more clearly identi ed 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. Articulatory (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 (Yavas, 2003) on perception. The current focus of interest in theoretical phonology, Optimality Theory, is treated in clinical phonology as though based on psycholinguistic 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 rst 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 clients speech output rst of all in terms of the data themselves. Systematic relationships (such as contrasts) should be allowed to emerge from the data. In addition, it is necessary to describe relationships between a clients realizations and target patterns, and to show whether any patterns uncovered are subject to variation (either contextual or stylistic). 4. A clinical phonological analysis should incorporate some kind of metric that allows an evaluation between the clients phonology and that of normal speakers, 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 problems that result from a disruption to the clients phonology, and those that result from disruptions to the motor planning or motor execution of an intact phonology (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).

Clinical Phonology 205 2. A clinical phonology should be able to distinguish between the source of a disordered 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 possibly also perception) process using a psycholinguistically adequate account. 4. A clinical phonology should inform therapeutic intervention based on a psycholinguistically adequate metric of the disorder and psycholinguistically relevant 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 psycholinguistically 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, nally, 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 re ection and debate about the nature of phonology, its representation and its prime (or smallest unit in phonological structure) has had very little in uence on clinical practice, with two exceptions: 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 speci c context of culture, namely that of speech-language pathology education and practice in the United States, since this is the context in which we have chie y 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

206 Martin J. Ball and Nicole Mller phonemes, and speech output. KLPA-2 uses the output of GFTA-2 in order to identify error patterns in a clients speech output, formulated in terms of phonological 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 individuals and across languages, in terms of mental operations affecting speech output. 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 eclectic and use whatever works. Thus speech sound errors are not theorized in clinical 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 childs 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 packaged 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 principles of analysis in clinical practice appear to be, rst 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 clients 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 patterns will show that a failure to theorize speech output problems, their origins and their effects, can obscure important patterns.

Clinical Phonology 207 The commonly used differentiation between distortion errors and substitution patterns arises out of the analysis of a clients 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 clients realization maps onto another unit in the target system (such as a clients production of [w] for /r/), an error is treated as a substitution, and by implication as an issue of the clients sound system. Indeed, the distinction between distortions and substitutions has been criticized by several authors. Grunwell (1985a, b, 1987, 1988) has commented on the inadequacy of these terms for the classi cation of speech errors. She has made it clear that often these two categories are insuf cient 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 Mller (2002) we may have a phonological cause with an articulatory effect, or an articulatory cause with a phonological 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 dif culty (e.g. ne motor control of the tongue apex in the case of apical []); one common realization ([w]) results in a phonological disruption to the clients 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 clinical 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 (psychological 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 processes, 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 super cially simple process is more complex. Gibbon has found that covert contrasts may well exist that are not obvious impressionistically. Instrumental techniques

208 Martin J. Ball and Nicole Mller (e.g. electropalatography) have shown that apparent mergers of alveolars with velars are often produced with signi cantly different tongue-palate contact patterns. So, what appears to the listener to be a phonological merger, is maintained by the speaker as phonetically different. Fricative stopping. Again, the effect of this pattern is to remove a whole category of target consonants from the childs 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 ne 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. 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 dif culty in the ne motor control of the tongue rims. Context sensitive voicing. At rst 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- nally). However, word nal devoicing is commonly encountered in natural language. Many Germanic languages, for example, only allow voiceless obstruents word- nally, and even in English there is considerable devoicing in word- nal 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 dif cult to maintain suf cient subglottal pressure for phonation as one nears the end of an utterance and air ow is expended. Conversely, it is easier to use phonation utterance initially when subglottal pressure is strong. 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 clients speech as singletons, but not as clusters. So, does this pattern derive from a phonology lacking the ability to have two (or more) consonants 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 alternative, motor explanation. Apart from the total absence of one member of a target 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 [ -] (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 hmans (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





Clinical Phonology 209 produce two consonants on the consonant parameter (as described by hman, 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 patterns 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 disorders that cannot be accounted for simply in articulatory terms. Ball and Mller (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 implicate 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 combines 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 de cits 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 distinction 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 revealing phenomena such as covert contrasts and undifferentiated gestures cast some doubt on these conclusions, suggesting that subtle phonetic dif culties 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

210 Martin J. Ball and Nicole Mller 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 signi cance of this point [constraint ranking] 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 topranked constraint presumably happens in a clients mental phonology. Yet, at the same time, the theory relies on an in nite number of possible pronunciations to be scanned by the evaluation (Eval) module, which as a model of speech production would surely result in no speech ever being produced. McCarthy (2002: 10) addresses this last point directly: how can Eval sort an in nite set of candidates in nite time . . . ? The error lies in asking how long Eval takes to execute. It is entirely appropriate to ask whether Eval . . . is well de ned, captures linguistically signi cant generalizations, and so on. But questions about execution time or other aspects of (neural) computation 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 articulatory or motor dif culties. 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 ne phonetic detail of much disordered speech. For example, the client who realizes word-initial English /r/ as a velar approximant (Mller et al., 2008) will t into no natural process, and will require an ad hoc constraint to characterize their speech. We turn in our nal 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 theoretical clinical phonology (Optimality Theory) have, in the course of clinical application, 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

Clinical Phonology 211 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 organization 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 constitute 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.

closed alveolar

narrow velar open lab wide


Figure 11.1 Gestural score for target alveolar in client showing velar fronting.

wide closed alveolar closed velar

(> khu)

narrow velar open lab


Figure 11.2 Gestural score for target velar sounding like alveolar in client showing velar fronting (showing covert contrast).

212 Martin J. Ball and Nicole Mller 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 adequate 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 rmly based within the traditions of phonology than many of those discussed there. Ball (2003) and Sosa and Bybee (2008) have both described how Bybees (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 phonology2 account deals with issues of storage of items, establishment of networks linking phonologically similar items and the emergence of units (such as contrastive units, or phonemes) from these networks. It posits highly redundant multiple storage, and thus does not seek the economy of description that traditional approaches do. We will brie y 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 Bybees (2001) model. She notes that experience affects representation; therefore high frequency forms and phrases have stronger representations in memory, and thus are resistant to analogic change. Low frequency forms are more dif cult 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 redundant 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

Clinical Phonology 213 storing units such as words, motor-articulatory plans to execute speech and acousticperceptual 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 generality. 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 at in terms of derivation, but multi-dimensional in terms of associations, as virtually all realizations are stored lexically. Nevertheless, Cognitive Phonology, as described by Bybee (2001) does recognize processes of simpli cation. 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 tted into Bybees (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, however, 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).

214 Martin J. Ball and Nicole Mller

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 Cognitive Phonology approach than in other popular clinical assessment types. We would need to describe data in terms of weak or non-existent lexical networks rst, leading to speci c 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). Bybees model does allow for different processes of simpli cation 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 insuf cient items are stored correctly in the lexicon, or if the categorization that creates connections does not occur. Of course, it may be dif cult 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 t well with our argument that much child disordered speech results from articulatory dif culties. 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 rst 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 (usually lenition). This affects high frequency forms most, as there is the greatest benet 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 dif culty (or a combination of the two).

11.7.4 Planning Remediation

In Bybees (2001) account of Cognitive Phonology, frequency of use is much more important than notions of contrast. Therefore, in remediation, minimal pairs

Clinical Phonology 215 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 speci c 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 unintelligible speech, one aspect of which was the use of double syllable onsets during syllable repetition tasks. These double onsets consisted of a prolonged rst consonant, a slight pause and then a second consonant leading into the remainder of the syllable. In some instances the rst and second onset consonants were the same, in other cases they were not; the rst 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: Repeat onset: [ - ] (target //) Disconnect nasality: [ - ] (target //) 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 nasality 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 realized correctly. Other English consonants were not included in this testing. In terms of gestural phonology, we can see all three of these patterns as examples of lenition, as they involve the removal of gestures and thereby the simpli cation of the gestural score. Figure 11.3 illustrates the gestural score for the syllable structure in [ ], where this simpli cation is clearly shown. This clients 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 dif culties with articulation which themselves may result in incorrect storage. The notion of density of storage suggested by Bybees work is also seen in the fact that the client could produce target consonants when prompted (i.e. she was

216 Martin J. Ball and Nicole Mller

crit alv

narrow velar open lab




Figure 11.3 Gestural score for [su]. (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 tting 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 nal clusters; addition of [] to words ending in /u/ and addition of a labial obstruent to words beginning with /r/. Following Bybees (2001) assertion that lenitions account for the majority of phonological processes, we assign the rst three of these patterns to that category. As noted immediately above, we do this under a gestural understanding of lenition, 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 (epenthetic vowels in clusters, and labial obstruents added to /r/ initial words), and for the addition of dark-l to /u/ nal words we see the addition of a tongue-tip gesture with the maintenance of the tongue body location through the back vowel into the

Clinical Phonology 217


narrow uvular closed labial open labial closed labial

narrow uvular open labial


Figure 11.4 Gestural scores for the nasal and denasal realizations of more. (From Ball et al., 2008.)

dark-l. It is dif cult to claim that these decoupling and reorganizing of gestures are also lenitions. As Bybee allows for different processes (e.g. fortitions) in acquisition, we also need such exibility 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 t 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 rst, 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 clients 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 psycholinguistic 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 rst descriptive criterion is theory independent; the requirement for working with naturalistic speech output is vital irrespective of analytic framework. The second point was the need to identify differences between target and realization. The gestural scores (as shown

218 Martin J. Ball and Nicole Mller in Figures 11.1 and 11.2) clearly show these differences, and are able to be combined to show patterns affecting similar sounds, thus satisfying the third criterion. The nal 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 envisage 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 rst 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 phonological 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, Bybees (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 phonology with Bybees 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 rst set of desiderata we listed above: those concerned with the description 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 pro table path for further research is via a cognitive phonology coupled to a gestural description.

1 It could be argued that using a tripartite division, Dodds (1995) delayed group would have a phonological disorder, her deviant consistent group could have a

Clinical Phonology 219

phonetic planning disorder, and her deviant inconsistent group an articulatory disorder. 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 Bybees (2001) model is the one explored here. 3 The following sections are based on Ball (2003).

Anderson, J. and Durand, J. (1986), Dependency phonology, in J. Durand, (ed.), Dependency and Non-Linear Phonology. London: Croom Helm, pp. 154. Anderson, J. and Durand, J. (eds) (1987), Explorations in Dependency Phonology. Dordrecht: Foris. Anderson, J. and Ewen, C. (1987), Principles of Dependency Phonology. Cambridge: Cambridge University Press. Archangeli, D. (1988), Aspects of underspeci cation theory, Phonology, 5, 183208. 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, 685702. Ball, M. J. (2003), Clinical applications of a cognitive phonology, Phoniatrics, Logopedics, Vocology, 28, 6369. 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, 447462. Ball, M. J. and Mller, N. (2002), The use of the terms phonetics and phonology in the description of disordered speech, Advances in Speech-Language Pathology, 4, 95108. Ball, M. J., Mller, N. and Damico, H. (2003), Double onset syllable structure in a speech disordered child, Advances in Speech-Language Pathology, 5, 3740. Ball, M. J., Rutter, B. and Code, C. (2008), Phonological analyses of a case of progressive speech degeneration, Asia-Paci c Journal of Speech, Language and Hearing, 11, 305312. 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 de cits? Clinical Linguistics and Phonetics, 10, 77101. Browman, C. P. and Goldstein, L. (1986), Towards an articulatory phonology, Phonology Yearbook, 3, 219252. Browman, C. P. and Goldstein, L. (1992), Articulatory phonology: an overview, Phonetica, 49, 155180. Bybee, J. (1985), Morphology: A Study of the Relation between Meaning and Form. Amsterdam: John Benjamins.

220 Martin J. Ball and Nicole Mller

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, 225252. 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. 15267. Code, C. F. S., Mller, N., Tree, J. T. and Ball, M. J. (2006), Syntactic impairments can emerge later: Progressive agrammatic aphasia and syntactic comprehension impairment, Aphasiology, 20, 10351058. Dinnsen, D. and Barlow, J. (1998), On the characterization of a chain shift in normal and delayed phonological acquisition, Journal of Child Language, 25, 6194. Dinnsen, D. and Gierut, J. (2008), The predictive power of optimality theory for phonological treatment, Asia-Paci c Journal of Speech, Language and Hearing, 11, 239249. 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. 126173. Gibbon, F. (1990), Lingual activity in two speech disordered childrens attempts to produce velar and alveolar stop consonants: Evidence from electropalatographic (EPG) data, British Journal of Disorders of Communication, 25, 329340. Gibbon, F. E. (1999), Undifferentiated lingual gestures in children with articulation / phonological disorders, Journal of Speech, Language, and Hearing Research, 42, 382397. Gibbon, F. E. (2007), Research and practice in developmental speech disorders, in M. C. Pennington (ed.), Phonology in Context. Basingstoke: Palgrave Macmillan, pp. 245273. Gibbon, F. E. and Scobbie, J. M. (1997), Covert contrasts in children with phonological disorder, Australian Communication Quarterly, Autumn 1997, 1316. 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. 5256. Gierut, J. A. (1990), Differential learning of phonological oppositions, Journal of Speech and Hearing Research, 33, 540549. Gierut, J. A. (1999), Syllable onsets: Clusters and adjuncts in acquisition, Journal of Speech, Language and Hearing Research, 42, 708726. 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.

Clinical Phonology 221

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, 165170. 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, 221252. 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. 457483. Grunwell, P. (1992b), Principled decision-making in the remediation of children with developmental phonological disorders, in P. Fletcher, and D. Hall (eds), Speci c Speech and Language Disorders in Children. London: Whurr, pp. 215240. Grunwell, P. (1997), Natural Phonology, in M. J. Ball, and R. D. Kent (eds), The New Phonologies. San Diego: Singular, pp. 3575. Harris, J. (1990), Segmental complexity and phonological government, Phonology, 7, 255300. 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, 6174. Harris, J. and Lindsey, G. (1995), The elements of phonological representation, in J. Durand and F. Katamba (eds), Frontiers of Phonology. London: Longmans, pp. 3479. Harris, J., and Lindsey, G. (2000), Vowel patterns in sound and mind, in N. BurtonRoberts, P. Carr and G. Docherty (eds), Phonological Knowledge, Conceptual and Empirical Issues. Oxford: Oxford University Press, pp. 185205. Hawkins, P. (1985), A tutorial comment of Harris and Cottam, British Journal of Disorders of Communication, 20, 7580. Hewlett, N. (1985), Phonological versus phonetic disorders: Some suggested modi cations to the current use of the distinction, British Journal of Disorders of Communication, 20, 155164. Kamhi, A. (2006), Treatment decisions for children with speech-sound disorders. Language, Speech, and Hearing Services in Schools, 37, 271279. Kaye, J., Lowenstamm, J. and Vergnaud, J.-R. (1985), The internal structure of phonological elements: A theory of charm and government, Phonology Yearbook, 2, 305328. Kaye, J., Lowenstamm, J. and Vergnaud, J.-R. (1990), Constituent structure and government in phonology, Phonology, 7, 193232. 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. 126. Kiparsky, P. (1982a), Lexical morphology and phonology, in I.-S. Yang (ed.), Linguistics in the Morning Calm. Seoul: Hanshin, pp. 391.

222 Martin J. Ball and Nicole Mller

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. 131175. Kiparsky, P. (1985), Some consequences of lexical phonology, Phonology Yearbook, 2, 83138. 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. 117145. McCarthy, J. (2002), A Thematic Guide to Optimality Theory. Cambridge: Cambridge University Press. Mller, N., Ball, M. J. and Rutter, B. (2008), Systemic phonology applied to an idiosyncratic case of /r/ disorder, Asia-Paci c Journal of Speech, Language and Hearing, 11, 269281. hman, S. (1966), Coarticulation in VCV utterances: Spectrographic measurements, Journal of the Acoustical Society of America, 39, 151188. Prince, A. and Smolensky, P. (1993), Optimality Theory: Constraint Interaction in Generative 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 assessment practices for children with suspected speech sound disorders: Results of a national survey. American Journal of Speech-Language Pathology, 16, 246259. Shriberg, L. D. and Kwiatkowski, J. (1982a), Phonological disorders II: A conceptual framework for management, Journal of Speech and Hearing Disorders, 47, 242256. Shriberg, L. D. and Kwiatkowski, J. (1982b), Phonological disorders III: A procedure for assessing severity of involvement, Journal of Speech and Hearing Disorders, 47, 256270. Sosa, A. V. and Bybee, J. (2008), A cognitive approach to clinical phonology, in M. J. Ball, M. Perkins, N. Mller 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. 80116. 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. 333363. Wheeler, D. and Touretsky, D. (1993), A connectionist implementation of cognitive phonology, in Goldsmith, J. (ed.), The Last Phonological Rule. Chicago: Chicago University Press, pp. 146172. Yavas M. (2003), Role of sonority in developing phonologies, Journal of Multilingual , Communication Disorders, 1, 7998.


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 elds of research. In this chapter we review some areas which are particularly relevant for the elds of applied linguistics, bilingualism and language acquisition. We begin by describing how British Sign Language (BSL) was rst documented. Next we look at how linguistic analysis has revealed the extent to which sign languages 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 rst language acquisition. In the nal section we look at the wider social context in which BSL is used including 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 rst century Mishnah (compilation of Jewish law) and earlier, the rst systematic observations of sign language use by deaf people in Britain date back to the seventeenth century. John Bulwers two books, Chirologia (1644) and Philocophus: Or the Deafe and Dumbe Mans Friend (1648), and George Dalgarnos 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 mans invention; yet you want not speeche; who have your whole body for a tongue, having a language more naturall and signi cant, 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 af nity 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 rst 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 arti cial 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 elaborates 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 reversion to earlier views about sign language. This shift can be understood in the context 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 language 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. (Bloom eld, 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 Bloom eldian structuralist model. Stokoe recognized that sign languages had an internal phonological structure comparable to that of spoken languages. Sign language researchers since 1960 have identi ed ve features of the sign languages of deaf communities throughout the world: They are complex natural human languages They are distinct from gesture There is no universal sign language They have their own grammars 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 af xation. 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 theory 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 explanations for the grammatical similarity among sign languages which still remain to be

226 Gary Morgan and Bencie Woll 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 continual 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 rst 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 subsequently became the rst teacher of the deaf in the United States. As soon as Clerc beheld this sight [the children at dinner] his face became animated; 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 communication (Maginn, F. (n.d.) Guide to the Silent Language of the Deaf. Belfast: Francis Maginn).

Sign Linguistics and Sign Bilingualism 227 approached them. He made signs and they answered him by signs. This unexpected 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 structure, with no clear evidence for recursiveness, and it is dif cult 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 signers own language; and the more experienced an individual is with IS, the more extensive the paraphrasing, suggesting increased awareness of the likely unintelligibility of structures taken from a speci c sign language.

12.5 Bimodal Bilingualism

Until recently, all studies of bilingualism were studies of individuals and communities 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: What are the rami cations of removing a major articulatory constraint on bilingual language production? 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 grammatical 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 identi ed both in behavioural and neuroimaging studies. For production, this is

228 Gary Morgan and Bencie Woll 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 nal common output path may impose a greater cognitive burden than selection of items within one language. For perception, similar constraints could apply the speakers 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 speci c cortical systems that may be somewhat independent of the languages themselves, and more involved in selection between speci c 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 manualgestural actions (signed languages) could in principle be processed simultaneously, 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 pro cient user of both languages, can therefore offer a unique insight into core language processes. Few studies have explored the implications of this situation, yet bimodal bilingualism is common. Hearing children of Deaf parents (sometimes called CODAs children of deaf adults) can acquire a sign language as a rst language. As adults, they have full access to at least two languages: a visualmanual 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 re ect a common conceptual source. Code blending re ects 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 CoDACoDA 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 community 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 organization of sign language have not taken into account that the deaf communities of

Sign Linguistics and Sign Bilingualism 229 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 impairment 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 crosstalk between the languages, and a different relationship between the lexicons of the bimodal bilinguals 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 language 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 simultaneously, 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 clari cation of core and modality-speci c 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 rst stems from

230 Gary Morgan and Bencie Woll a scienti c interest in language development and modality. The second is a more practical reason. Since the vast majority of deaf children are born to hearing parents, 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 communicate 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 impoverished exposure to a rst 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 childrens 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 speci c 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 rst words and word combinations 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 pro cient 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 speci c 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 signing, 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 uent 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. Adultchild signing is slower, more exaggerated with larger movement and tends to have more repetition than adultadult 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 af xes. This means that children are required to pay attention visually

Sign Linguistics and Sign Bilingualism 231 to adults signing and conversely adults have to time their signing to moments when children are looking at them. These two constraints, (1) looking at the communication partner and (2) timing the input to the child at moments when the child is looking, are not necessary in spoken language acquisition. Hearing children 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 speci c characteristics of child directed signing. (Harris, 2001). One common nding from several studies of early child-parent sign language interaction is that the number of signs addressed to young children by uent signing adults using ASL, BSL and Sign Language of the Netherlands (SLN) is considerably less than that in hearing parenthearing child interaction in the same community and also less than in adultadult 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 uent adult signers there is no evidence that this slows language development 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 childs gaze is at the adult signer, and it may contain child directed modi cations to the grammar such as the dropping of agreement or aspectual in ections 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 hearing childrens 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 nger 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 st has thumb and little nger protruding) (Sutton-Spence and Woll, 1999). This direct link between meaning and form may in uence childrens development of sign language as rst languages. Before discussing the data on rst 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 objects 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 1224-month olds as equally able to see iconic links as adults who are learning sign as a second language?

232 Gary Morgan and Bencie Woll 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 childrens rst signs are closely linked to iconic properties of the referent. Childrens rst signs are more likely to be associated with the same sets of semantic categories evident in childrens early speech e.g. signs for people, animals and food (Anderson and Reilly, 2002). In a study of rst language acquisition of BSL from age 1;10 3;0, Morgan, Barriere and Woll (2003) reported that from age 1;10 onwards the majority of signs that were being acquired by the child were non-iconic. Children who acquire a sign language as their rst 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 during 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 childrens rst signs appeared 13 months prior to the rst words of hearing peers, at around 79 months of age or earlier. Other research reported that rst signs emerge closer to the magic 12 months and at the same age as hearing childrens rst 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 rst symbolic and intentionally communicative use of signs has not replicated the ndings from the early studies. Across different studies, rst 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 rst 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 childs 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 rst hand-internal movements to appear in manual babbling in ASL (e.g. Cheek et al., 2001), hearing childrens gestures (Volterra et al., 2005) and childrens rst signs (Cheek et al., 2001). Marschark (2002) suggests that deaf parents may pay more direct attention to their childrens manual productions than is normally the case for hearing childadult interaction thus spotting gestures that t a particular communicative function earlier. Further as Meier and Newport (1990) suggested, the fact that all of a signs component parts are visible to parents and researchers compared with signi cantly less of childrens rst spoken words may lead some to see partial signs earlier.

Sign Linguistics and Sign Bilingualism 233 Importantly when Meier and Newport (1990) reviewed the studies of early signing 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 childrens 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 nd the acquisition of these abstract principles any easier than children acquiring a spoken language (see also Meier, 1987). The dif culty in deciding if hand movements of young children are rst signs or gesturing may explain some of the claims made in the media about the developmental 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 normally 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 rst language produce their rst signs slightly earlier than babies produce their rst spoken words. Several explanations have been put forward for this reported sign advantage. One suggestion is the differing maturation 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 themselves 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. ngers to lips for eat, arms out straight like wings for airplane) http://www.babysigns.com.sg/FAQ.html. If speci c gestures and sounds are reinforced, babies will develop communication 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 gestures than speech sounds but it is a transitory phase. By the time children have

234 Gary Morgan and Bencie Woll started to produce their rst 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 follow. 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 uency. 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 monolingual oral schools while 1,114 (4.45 per cent) attended schools with a sign bilingual 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 making 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 40177 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 education systems that report using sign language (Greenberg and Kusch, 1987; Meadowet 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

Sign Linguistics and Sign Bilingualism 235 to develop deaf childrens metalinguistic awareness of the differences and similarities 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 representation of BSL in the media. As recently as 1989, the following review ridiculing the sign language interpreted version of the Queens Christmas message appeared in the Guardian: The funniest thing I saw was the Queens 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 educators 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 19952006. The cumulative total of people with at least a basic knowledge of BSL are now over 190,000 (a signi cantly 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

236 Gary Morgan and Bencie Woll

18000 16000 14000 12000 10000 8000 6000 4000 2000 0 1995 1996 1997 1999 2000 2001 2002 2003



BSLStage 2

BSLStage 1

Figure 12.2 Students taking BSL exams (Woll, 2001. Additional data provided by Signature: the Council for the Advancement of Communication with Deaf People).

funding for sign language interpretation for undergraduate and postgraduate students, and the Disability Discrimination Act requires the provision of sign language interpretation by rms 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 quali ed sign language interpreters over the past 24 years. The recent rapid rise in the number of quali ed interpreters re ects shortterm projects that have sought to move trainees to quali ed 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 Of cial Recognition of BSL

The British Deaf community campaigned for many years for of cial recognition of BSL by the UK government which nally 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

Sign Linguistics and Sign Bilingualism 237

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 quali ed 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- rst century, there are two contrasting futures: on the one hand, there are pressures, such as the decrease in opportunities for Deaf children to use BSL with their peers as a result of the move to mainstream education, and the possible decrease in the Deaf population as a result of medical intervention and advances in genetics; on the other hand, increased interest 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 illuminating the relationship of language form to communication channel. In the next ten

238 Gary Morgan and Bencie Woll years we will see more sophisticated uses of brain imaging techniques to explore brain plasticity, particularly in situations of late rst and second language acquisition, and in contrasting and comparing bimodal and unimodal bilingualism. The growth of interest in gestural communication and in the evolution of human language have both contributed to and been enriched by sign language research, and these elds 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 achievements 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 schoolleavers. New research into language development and the learning of literacy and numeracy skills needs to feed into this educational policy review.

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. 171188. 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. 83119. Barlow, I. H. and Brill, R. G. (1975), An evaluation of reading and academic achievement levels of 16 graduating classes of the California School for the Deaf, Riverside, Volta Review, 77, 255266. 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 Bloom eld, L. (1933), Language. New York: Holt Rinehart Winston. Bonvillian, J. and Siedlecki, T. (1996), Young childrens acquisition of the location aspect of ASL signs: parental report ndings, Journal of Communication Disorders, 29, 1325. Bonvillian, J., Orlansky, M. and Novack, L. (1993), Developmental milestones: Sign language acquisition and motor development, Child Development, 54, 14351445. 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 Mans 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), 292323. 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, 663673. Fischer, S. D. (1978), Sign language and creoles, in P. Siple (ed.), Understanding Language through Sign Language Research. New York: Academic Press.

Sign Linguistics and Sign Bilingualism 239

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. 197218. Greenberg, M. and Kusch, 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, 95129. Harris, M. (2001), Its 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), 177185. 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, 95103. 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), 3347. 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, 235251. Kantor, R. (1982), Communicative interactions: Mother modi cation and child acquisition of American Sign Language, Sign Language Studies, 11, (36), 233282. Kegl, J., Senghas, A., and Coppola, M. (1999), Creation through contact: sign language emergence and sign language change in Nicaragua, in M deGraff (ed.), Language Creation and Language Change. Cambridge, MA: The MIT Press, pp. 179237. 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), International Studies on Sign Language and Communication of the Deaf 10. Hamburg: Signum, pp. 200211. 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 Underlying British Sign Language and Audio-Visual English Processing in Native Users. Brain 125, 15831593. Mahshie, S. (1995), Educating Deaf Children Bilingually. Washington, DC: Gallaudet University 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, 128. 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 Neuropsychology, 22, 719736.

240 Gary Morgan and Bencie Woll

Masataka, N. (1996), Perception of motherese in a signed language by 6-month-old deaf infants, Developmental Psychology, 32, 874879. 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. 291296. 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, 454468. Meier, R. P. (1987), Elicited imitation of verb agreement in American Sign Language: iconically or morphologically determined? Journal of Memory and Language 26. (3), 362376. 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), 123. 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. 314343. 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. 5766. 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. 881938. 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 childrens rst 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. 2558. 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, 261286. 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, 283296. Rosenstock, R. (2008), The role of iconicity in international sign, Sign Language Studies, 8, 131159. Schick, B. S. (1990), The effects of morphosyntactic structure on the acquisition of classi er predicates in ASL, in C. Lucas (ed.), Sign Language Research: Theoretical Issues. Washington, DC: Gallaudet University Press, pp. 358374.

Sign Linguistics and Sign Bilingualism 241

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. 219231. 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. 344373. Spencer, P. E. (1993), The expressive communication of hearing mothers and deaf infants, American Annals of the Deaf, 138, 275283. 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. 333352. 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, Montral, 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 emergence and development of language, in Tomasello, M., Slobin, D. (eds), Elizabeth Bates: A Festschrift. Mahwah, NJ: Lawrence Erlbaum Associates, pp. 340. 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. 7992. 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 Association for Applied linguistics), in association with Continuum, pp. 6580.

This page intentionally left blank

Personal Name Index

Archangeli, D. 203, 209 Ball, M. 8, 202, 207, 209, 213, 215, 217, 219 Bassnett, S. 146, 161 Bhatia, V.K. 165, 166 Block, D. 2, 11 Bourdieu, P. 2, 6, 11 Browman, C. 204, 211 Bush, G.W. 52, 144, 157, 158, 159, 160 Bybee, J. 212, 213, 214, 215, 216, 217, 218, 219 Cameron, D. 4, 11, 15, 22 Chomsky, N. 2, 28, 98, 189, 193, 202, 203 Code, C. 209, 216, 217 Coulmas, F. 7, 34, 38 Coulthard, M. 178, 179 Crystal, D. 4, 48, 115, 166 de Saussure, F. 28, 100 Dinnsen, D. 204, 210 Dodd, B. 209, 218 Dunbar, R. 85, 86, 87, 90, 91, 92 Eades, D. 169, 171, 176, 179 En eld, N. 7, 83, 85, 86, 88, 89, 90, 91 Ferguson, C. 70, 76 Fillmore, C. 101, 165 Fishman, J. 36, 66, 68 Foucault, M. 2, 19 Gibbons, J. 8, 164, 168, 170, 171 Gierut, J. 208, 210, 215 Goffman, E. 85, 89, 91, 92, 109 Goldstein, L. 204, 211 Goodwin, C. 85, 88 Greenberg, J. 36, 234 Grice, P. 31, 89 Grin, F. 11, 16, 32, 35, 36 Grunwell, P. 202, 207 Gu, Y. 7, 107, 109, 110, 114, 115, 118, 119 Gunnarsson, B. 7, 123, 124, 125, 127, 128, 129, 130, 132, 134 Halliday, M. 165, 166 Heller, M. 10, 11, 15, 16 Hinde, R. 82, 85, 86, 87, 92 Killingley, D. 76, 77 Kockelman, P. 83, 84, 85, 86, 88, 89, 92 Koskinen, K. 145, 147, 150, 154 Levinson, S. 83, 86, 88, 89 Lorch, M. 8, 186, 189, 194, 195, 196, 197 Meier, R. 225, 230, 232, 235 Morgan, G. 8, 230, 232, 234 Muller, N. 8, 207, 209, 210 Nettle, D. 36, 47, 48, 60 OBarr, W. 169, 187 Paradis, M. 188, 189, 190, 191, 192, 193, 195 Pavlenko, A. 7, 10 Piller, I. 7, 10, 16, 19, 20 Pitres, A. 189, 190, 195, 196 Robinson, C. 55, 56, 58 Robinson, C.D.W. 32, 33 Romaine, S. 7, 36, 47, 48, 56, 58, 59, 60, 61 Rubinstein, A. 29, 30 Sacks, H. 83, 86, 91 Sawyer, J. 70, 75 Schaffner, C. 8, 145, 146, 161 Shuy, R. 172, 177 Spolsky, B. 7, 65, 66, 68, 71, 72, 73 Tomasello, M. 84, 85, 87 Woll, B. 8, 226, 230, 231, 232, 234, 235, 236

Subject Index
activity zone 105, 1078, 114 Africa 14, 16, 40, 46, 48, 51, 54, 55, 65, 69, 71, 74, 75, 76, 133, 170 age 21, 54, 176, 185, 186, 191, 206, 232, 234 aphasia 185, 187, 189, 190, 191, 195, 229 Arabic 48, 65, 69, 746, 77, 78, 131, 146 Aramaic 66, 67, 69, 75, 76 Asia 14, 16, 74, 75, 76, 77, 133, 164 Australia 20, 22, 49, 50, 54, 61, 73, 167 Bible 65, 66, 69, 70, 71, 72 bilingual aphasia 189, 190, 191, 192, 194, 195, 229 bilingualism 2, 4, 8, 12, 13, 14, 22, 78, 130, 133, 172, 191, 223ff Brazil 50, 52, 128 Britain 124, 125, 143, 149, 167, 223 Canada 15, 32, 49, 50, 53, 60, 61, 189 child-rearing 12, 21, 22 child speech disorder 202, 209, 210 China 37, 48, 50, 51, 52, 72, 99, 103, 104, 107, 108, 111, 114, 164 Chinese 15, 33, 34, 35, 38, 48, 52, 98, 99, 102, 103, 110, 114, 115, 146, 167 choice 29, 31, 33, 34, 39, 41, 65, 66, 67, 69, 73, 77, 127, 128, 129, 130, 133, 151, 178, 215 Christianity 66, 69, 70, 71, 73, 74 clinical data 202, 204, 205, 209, 210, 212, 214 cognitive process 8, 84, 85, 223 colonialism 37, 46, 47 common language 36, 38, 39, 40, 66 complexity 33, 39, 84, 122, 150, 161, 167, 168, 170, 205 Congo 49, 50, 51, 71, 72 corporate language 122, 130, 131 covert contrast 202, 207, 209, 211, 218 Critical Discourse Analysis (CDA) 6, 122, 161, 169 cross-linguistic 38, 189, 226 culture 7, 35, 47, 57, 59, 60, 72, 75, 83ff, 99, 123, 124, 125, 126, 127, 129, 132, 134, 136, 142, 149, 151, 160, 161, 171, 172, 173, 189, 205, 237 deaf community 89, 168, 223ff development 2, 7, 16, 36, 37, 40, 41, 46ff, 70, 71, 72, 74, 76, 77, 85, 86, 88, 108, 135, 186, 187, 188, 193, 194, 195, 196, 203, 205, 208, 214, 218, 224, 230, 231, 232, 233, 238 diversity 3, 7, 36, 40, 46ff, 122, 123, 127, 129, 133, 134, 135, 137, 138 economic theory 28, 29, 35, 38 economy 7, 10, 11, 12, 13, 16, 22, 28ff, 122, 123, 129, 135, 136, 137, 176 education 2, 8, 11, 12, 13, 14, 32, 35, 53, 56, 57, 60, 61, 71, 72, 73, 104, 106, 113, 114, 133, 143, 171, 172, 185, 187, 189, 206, 223, 228, 234, 237, 238 Egyptian 40, 69, 73 English 10, 11, 13, 14, 15, 17, 18, 20, 22, 34, 37, 38, 40, 41, 47, 48, 57, 58, 69, 71, 75, 78, 98 122, 123, 124, 127, 128, 129, 130, 131, 132, 133, 141, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 158, 159, 160, 162, 164, 166, 167, 168, 170, 171, 172, 173, 177, 178, 185, 189, 193, 195, 202, 205, 207, 208, 210, 213, 229, 235 Europe 2, 3, 15, 16, 18, 19, 28, 34, 35, 37, 38, 48, 59, 66, 67, 68, 70, 71, 74, 75, 90, 124, 129, 133, 143, 144, 145, 146, 147, 148, 149, 150, 151, 153, 154, 157, 164, 167, 172, 177, 229, 236 external communication 122, 126, 130, 135 face-to-face interaction 10, 92, 102, 106, 107, 109, 118, 166 forensic linguistics 1, 8, 176, 179 formality 161, 165, 178 French 2, 3, 11, 13, 19, 20, 32, 38, 55, 56, 67, 71, 78, 98, 128, 142, 144, 146, 147, 148, 149, 150, 153, 154, 155, 156, 167, 231

Subject Index
GDP (gross domestic product) 34, 36, 50, 57, 59 gender 7, 10ff, 36, 169, 176, 188 genre 21, 135, 145, 161, 165, 166, 172, 178 German 13, 15, 22, 38, 67, 38, 77, 78, 124, 125, 128, 131, 134, 142, 143, 144, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 166 gestural phonology 204, 211, 214, 215, 218 global 7, 11, 16, 19, 40, 41, 46, 47, 48, 49, 51, 52, 57, 58, 79, 122, 123, 127, 128, 129, 135, 136, 137, 146 Greek 66, 67, 68, 69, 70, 146, 171 Hebrew 18, 65, 66, 67, 68, 72, 77 Hindi 48, 71, 75, 77, 146 human rights 58, 59, 61 identity 10, 19, 34, 35, 36, 46, 86, 110, 112, 113, 124, 134, 147, 191 idiolect 104, 105, 106, 114, 115 immigrant 14, 17, 18, 19, 21, 33, 66, 67, 73, 128, 129, 131, 133, 134, 143, 191 India 50, 51, 53, 54, 71, 72, 74, 75, 76, 77, 170 indigenous 7, 12, 13, 46, 47, 53, 54, 56, 57, 58, 59, 60, 61, 71, 72, 176 inequality 16, 38, 47, 51, 52 interpersonal 91, 104, 114, 165, 168 interpreter 15, 39, 76, 89, 157, 170, 171, 235, 236, 237 interpreting 143, 146, 156, 157, 170, 171, 236 intervention 56, 74, 155, 173, 175, 202, 205, 206, 237 Islam 65, 66, 74, 75, 76, 79, 162, 166 Israel 67, 68, 79, 143, 144, 162 Japanese 20, 21, 34, 38, 48, 59, 128, 164, 167 Judaism 65, 66, 68, 69, 72, 73, 74 language acquisition 1, 2, 4, 5, 33, 195, 223, 229, 230, 231, 232, 238 language choice 33, 34, 35, 39, 65, 77, 128, 130 language shift 13, 37, 60, 172 Latin 65, 69, 70, 73, 167, 168, 175


Latin America 12, 14, 54, 70, 76, 164 law 5, 58, 67, 68, 72, 95, 75, 108, 110, 117, 147, 164ff, 223, 257 Lebanon 75, 78, 134 lingua franca 13, 72, 74, 123, 129, 130, 131, 133, 137 local 11, 33, 38, 55, 56, 65, 67, 70, 71, 72, 73, 75, 76, 78, 86, 87, 89, 90, 92, 105, 106, 107, 111, 123, 124, 125, 127, 128, 129, 132, 133, 135, 137, 169 marginalization 56, 129, 134 marketing 8, 22, 123, 125, 126 MDG (Millennium Development Goal) 19, 56, 61 media 2, 16, 38, 47, 105, 115, 116, 117, 118, 142, 143, 144, 145, 156, 163, 173, 235 Mexico 12, 50, 52, 54, 78 migration 10, 11, 14, 17, 19, 56, 67, 114, 186 minority 12, 13, 14, 15, 16, 18, 21, 22, 35, 36, 40, 55, 56, 61, 67, 105, 128, 171, 172, 190, 230, 234, 236, 237 missionary 65, 70, 71, 72, 76 mobility 108, 114, 118, 122, 135, 136 modernization 4, 58, 76, 169 Mongolia 52, 104, 107, 108 monolingualism 2, 14, 15, 47, 143 motor planning 204, 207, 213, 217 multilingualism 1, 2, 4, 10ff, 34, 36, 37, 122, 129, 130, 189, 192 multimodality 105, 109, 110, 111, 112, 113, 135 native 18, 49, 53, 54, 58, 60, 103, 111, 131, 133, 178, 190, 195, 225, 232, 234 network 5, 10, 20, 35, 39, 41, 46, 83, 112, 113, 122, 136, 137, 147, 191, 212, 213, 214, 215, 229 New Zealand 61, 70, 71, 73, 138 Nigeria 13, 49, 50, 51, 53 norm 88, 89, 90, 91, 146, 162 Optimality Theory 204, 209, 210 organization 7, 8, 20, 25, 31, 38, 39, 41, 66, 72, 74, 84, 85, 122ff, 145, 146, 151, 185, 187, 188, 189, 192, 195, 197, 203, 208, 211, 217, 223, 228

246 Subject Index

Papua New Guinea 46, 50, 53 perception 57, 59, 84, 204, 205, 212, 214, 228 phonological 8, 79, 193, 202ff, 223, 225 Policy 6, 7, 33, 34, 39, 40, 41, 46, 56, 58, 60, 61, 65, 66, 68, 69, 70, 71, 72, 73, 74, 76, 77, 110, 127, 129, 134, 137, 145, 146, 147, 148, 152, 175, 234, 238 political 8, 36, 37, 40, 58, 61, 73, 85, 98, 107, 118, 142ff Portuguese 48, 70, 128, 156, 157 poverty 7, 30, 36, 46ff prayer 65, 66, 67, 69, 78 problem-solving 1, 2, 90, 92 psycholinguistics 116, 191, 203, 204, 205, 210, 212, 213, 217, 218 Quran 74, 75, 76, 78 rhetoric 16, 29, 30, 31, 61 Russian 18, 48, 52, 70, 73, 78, 131, 144, 146 school 5, 15, 20, 21, 37, 40, 50, 53, 55, 56, 57, 58, 60, 68, 70, 71, 72, 75, 76, 79, 104, 114, 133, 173, 185, 187, 191, 203, 224, 226, 228, 234, 238 situatedness 105, 107, 110, 111, 112, 118 social relations 6, 7, 83, 86, 91 sociolinguistics 66, 115, 122, 134, 191 South-East Asia 18, 75, 164 Spanish 12, 13, 14, 48, 58, 70, 78, 131, 142, 146, 147, 148, 154, 171, 177 speech disorder 195, 202, 209, 210, 213 speech production 8, 202, 203, 104, 105, 120, 212, 218 spoken discourse 125, 132, 178 Sudan 50, 51, 75, 79 translation 3, 8, 38, 39, 65, 66, 67, 68, 69, 70, 71, 72, 74, 75, 76, 78, 142ff, 170, 171 trust 31, 38, 41, 86 United Nations (UN) 58, 61, 146 United States (US) 14, 19, 32, 46, 49, 50, 51, 52, 54, 59, 61, 67, 68, 73, 78, 103, 157, 158, 167, 170, 172, 173, 174, 176, 184, 192, 205, 226 utility 32, 33, 34, 35, 36, 37, 39, 40 vernacular 65, 66, 67, 68, 69, 70, 71, 72, 73, 75, 77, 78 website 106, 127, 128, 130, 146, 151, 152, 154, 155, 156, 157, 158, 159, 160, 233 well-being 54, 56, 57, 60 workplace 111, 114, 118, 122, 123, 124, 129, 132, 134, 135, 136, 137, 138 writing 29, 40, 60, 75, 85, 99, 110, 117, 123, 124, 125, 127, 131, 135, 178, 186, 196, 206 Yiddish 67, 68, 69, 79