Applied Linguistics, volume 31, issue 2 , 2010

APPLIN-31(2)Cover.qxd

3/30/10

5:54 PM

Page 1

APPLI ED LI NGUISTICS

ISSN 0142-6001 (PRINT) ISSN 1477-450X (ONLINE)

Applied Linguistics

Volume 31 Number 2 May 2010

Volume 31 Number 2 May 2010

Published in cooperation with AAAL American Association for Applied Linguistics AILA International Association of Applied Linguistics BAAL British Association for Applied Linguistics

OXFORD

www.applij.oxfordjournals.org

APPLIN-31(2)Cover.qxd

3/30/10

5:54 PM

Page 2

EDITORS

NOTES TO CONTRIBUTORS

Ken Hyland, Director, Centre for Applied English Studies, KK Leung Building, The University of Hong Kong, Pokfulam Road, Hong Kong Jane Zuengler, Nancy C. Hoefs Professor of English, University of Wisconsin-Madison, 6103 Helen C. White 600 North Park Street Madison, WI, 53706 USA Assistant to Jane Zuengler: Heather Carroll, University of Wisconsin-Madison

Articles submitted to Applied Linguistics should represent outstanding scholarship and make original contributions to the field. The Editors will assume that an article submitted for their consideration has not previously been published and is not being considered for publication elsewhere, either in the submitted form or in a modified version. Articles must be written in English and not include libelous or defamatory material. Manuscripts accepted for publication must not exceed 8,000 words including all material for publication in the print version of the article, except for the abstract, which should be no longer than 175 words. Additional material can be made available in the online version of the article. Such additions will be indexed in the print copy.

REVIEWS AND FORUM EDITOR Stef Slembrouck, Professor of English Linguistics and Discourse Analysis, Universiteit Gent, Vakgroep Engels, Rozier 44, B-9000 Gent, Belgium. <[email protected]> Assistant to Stef Slembrouck: Tine Defour, Universiteit Gent

ADVISORY BOARD Guy Cook, British Association for Applied Linguistics Aneta Pavlenko, American Association for Applied Linguistics Martin Bygate, International Association for Applied Linguistics Huw Price, Oxford University Press

EDITORIAL PANEL Karin Aronsson, Linko¨ping University David Block, London University Institute of Education Jan Blommaert, University of Jyva¨skyla¨ Deborah Cameron, University of Oxford Lynne Cameron, Open University (BAAL Representative) Tracey Derwing, University of Alberta Zolta´n Do¨rnyei, University of Nottingham Patricia Duff, University of British Columbia Diana Eades, University of New England, Australia ZhaoHong Han, Columbia University (AAAL representative) Gabriele Kasper, University of Hawai’i at Manoa Claire Kramsch, University of California at Berkeley Angel Lin, City University Hong Kong Janet Maybin, Open University, UK Tim McNamara, University of Melbourne Junko Mori, University of Wisconsin-Madison Greg Myers, Lancaster University Susanne Niemeier, University Koblenz-Landau (AILA Representative) Lourdes Ortega, University of Hawai’i at Manoa Alastair Pennycook, University of Technology, Sydney Ben Rampton, King’s College, University of London Steven Ross, Kwansei Gakuin University Alison Sealey, University of Birmingham Antonella Sorace, University of Edinburgh Lionel Wee, National University of Singapore Applied Linguistics is published five times a year in February, May, July, September and December by Oxford University Press, Oxford, UK. Annual subscription price is £254/US$457/E381. Applied Linguistics is distributed by Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. Periodicals postage paid at Rahway, NJ and at additional entry points. US Postmaster: send address changes to Applied Linguistics (ISSN 0142-6001), c/o Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. # Oxford University Press 2010 All rights reserved; no part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without prior written permission of the Publishers, or a licence permitting restricted copying issued in the UK by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1P 9HE, or in the USA by the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. Typeset by Glyph International, Bangalore, India Printed by Bell and Bain Ltd, Glasgow

Applied Linguistics operates a double-blind peer review process. To facilitate this process, authors are requested to ensure that all submissions, whether first or revised versions, are anonymized. Authors’ names and institutional affiliations should appear only on a detachable cover sheet. Submitted manuscripts will not normally be returned. Forum pieces are usually reviewed by the journal Editors and are not sent for external review. Items for the Forum section are normally 2,000 words long. Contributions to the Forum section and offers to review book publications should be addressed to the Forum and Reviews Editor. For more detailed guidelines, see our website http://www.oxfordjournals.org/applij/for_authors/index.html

PROOFS Proofs will be sent to the author for correction, and should be returned to Oxford University Press by the deadline given.

OFFPRINTS On publication of the relevant issue, if a completed offprint form has been received stating gratis offprints are requested, 25 offprints of an article, forum piece or book review will be sent to the authors free of charge. Orders from the UK will be subject to a 17.5% VAT charge. For orders from elsewhere in the EU you or your institution should account for VAT by way of a reverse charge. Please provide us with your or your institution’s VAT number.

COPYRIGHT Acceptance of an author’s copyright material is on the understanding that it has been assigned to the Oxford University Press subject to the following conditions. Authors are free to use their articles in subsequent publications written or edited by themselves, provided that acknowledgement is made of Applied Linguistics as the place of original publication. Except for brief extracts the Oxford University Press will not give permission to a third party to reproduce material from an article unless two months have elapsed without response from the authors after the relevant application has been made to them. It is the responsibility of the author to obtain permission to reproduce extracts, figures, or tables from other works.

Applied Linguistics Journal online The full text of Applied Linguistics is available online to journal subscribers. Online access has a number of advantages: . quality PDFs ensure articles look the same as the print original and are easy to print out . access is easy—all you need is your subscription number or institutional IP address (see below) . online access is available ahead of print publication—so view while you await your print version! . access the text wherever you are (or from any part of your institution network if you have a library subscription) . perform searches by word or author across the full text of the articles of any part of the journal . download articles whenever you choose—you will be able to access past online issues as long as you have a current subscription . free sample copy available online . fully searchable abstracts/titles going back to volume 1 . Table of Contents email alerting service. The print version will continue to be available as previously. Institutions may choose to subscribe to the print edition only, online only, or both. Individual subscribers automatically receive both.

CONTRIBUTORS There is no need for contributors to format their articles any differently; online files are produced automatically from the final page proofs of the journal. However, if you know that an item in your list of references is available online, please supply the URL. If you have your own website, you are welcome to include the URL with your contact address in your biodata.

ADVANCE ACCESS Applied Linguistics now has Advance Access articles. These are papers that have been copyedited and typeset but not yet paginated for inclusion in an issue of the journal. More information, including how to cite Advance Access papers, can be found online at http://www.applij.oxfordjournals.org.

Applied Linguistics Subscription Information

A subscription to Applied Linguistics comprises 5 issues. Annual Subscription Rate (Volume 31, 5 issues, 2010) Institutional. Print edition and site-wide online access: £254.00/US$457.00/E381.00; Print edition only: £233.00/US$419.00/E350.00; Site-wide online access only: £212.00/US$382.00/E318.00.

Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, UK. Email: [email protected]. Tel (and answerphone outside normal working hours): +44 (0)1865 353907. Fax: + 44 (0)1865 353485. In Japan, please contact: Journals Customer Services, Oxford Journals, Oxford University Press, Tokyo, 4-5-10-8F Shiba, Minato-ku, Tokyo 108-8386, Japan. Tel: þ81 3 5444 5858. Fax: þ81 3 3454 2929.

Personal. Print edition and individual online access: £82.00/US$164.00/E123.00.

subscribe to applied linguistics

Please note: US$ rate applies to US & Canada, Euros applies to Europe, UK£ applies to UK and Rest of World.

For new subscriptions and recent single issues only. Current subscribers will automatically receive a renewal form.

Prices include postage by surface mail, or for subscribers in the USA and Canada by airfreight, or in India, Japan, Australia and New Zealand, by Air Speeded Post. Airmail rates are available on request. There are other subscription rates available for members of AAAL, BAAL, AILA, and LSA, for a complete listing please visit www.applij.oxfordjournals.org/subscriptions. Full prepayment, in the correct currency, is required for all orders. Orders are regarded as firm and payments are not refundable. Subscriptions are accepted and entered on a complete volume basis. Claims cannot be considered more than FOUR months after publication or date of order, whichever is later. All subscriptions in Canada are subject to GST. Subscriptions in the EU may be subject to European VAT. If registered, please supply details to avoid unnecessary charges. For subscriptions that include online versions, a proportion of the subscription price may be subject to UK VAT. Personal rate subscriptions are only available if payment is made by personal cheque or credit card and delivery is to a private address. The current year and two previous years’ issues are available from Oxford Journals. Previous volumes can be obtained from the Periodicals Service Company at http://www.periodicals. com/oxford.html or Periodicals Service Company, 11 Main Street, Germantown, NY 12526, USA. Email: [email protected]. Tel: +1 (518) 537 4700. Fax: +1 (518) 537 5899. For further information, please contact: Journals Customer Service Department,

Please complete the form below and return it to: Journal Customer Service Department (please see above). Please record my subscription to Applied Linguistics, starting with Volume__________ (Subscriptions start with the March issue and can be accepted for complete volumes only.) Please send me the following single issue(s) Volume_________ Issue_________ Name (BLOCK CAPITALS please) _________________________________________ Adresss __________________________________ _________________________________________ _________________________________________ City _____________________________________ Country _________________________________ Postcode _________________________________ I enclose the correct payment of (see rates above): £/US/E __________________________________ Please debit my credit card: American Express / Mastercard / Visa (delete as appropriate) Card number: __|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__|__ Expiry date: |__|__|__| Signature ________________________________ œ Please tick this box if you do NOT wish to receive details of related products and services of OUP and other companies that we think may be of interest.

Aims Applied Linguistics publishes research into language with relevance to real-world problems. The journal is keen to help make connections between fields, theories, research methods, and scholarly discourses, and welcomes contributions which critically reflect on current practices in applied linguistic research. It promotes scholarly and scientific discussion of issues that unite or divide scholars in applied linguistics. It is less interested in the ad hoc solution of particular problems and more interested in the handling of problems in a principled way by reference to theoretical studies. Applied linguistics is viewed not only as the relation between theory and practice, but also as the study of language and language-related problems in specific situations in which people use and learn languages. Within this framework the journal welcomes contributions in such areas of current enquiry as: bilingualism and multilingualism; computer-mediated communication; conversation analysis; corpus linguistics; critical discourse analysis; deaf linguistics; discourse analysis and pragmatics; first and additional language learning, teaching, and use; forensic linguistics; language assessment; language planning and policies; language for special purposes; lexicography; literacies; multimodal communication; rhetoric and stylistics; and translation. The journal welcomes both reports of original research and conceptual articles. The Journal’s Forum section is intended to enhance debate between authors and the wider community of applied linguists (see Editorial in 22/1) and affords a quicker turnaround time for short pieces. Forum pieces are typically responses to a published article, a shorter research note or report, or a commentary on research issues or professional practices. The Journal also contains a Reviews section. Applied Linguistics is covered by the following abstracting/indexing services: Bibliographie Linguistique/Linguistic Bibliography, BLonline, British Education Index, Current Index to Journals in Education, ERIC (Education Resources Information Centre), International Bibliography of the Social Sciences, ISI: Social Sciences Citation Index, Research Alert, Current Contents/Social and Behavioral Sciences, Social Scisearch, Sociological Abstracts: Language and Linguistics Behaviour Abstracts, Language Teaching, MLA Directory of Periodicals, MLA International Bibliography, PsycINFO, Sociological Abstracts, Zeitschrift fu¨r Germanistische Linguistik.

ADVERTISING Inquiries about advertising should be sent to Linda Hann, Oxford Journals Advertising, 60 Upper Broadmoor Road, Crowthorne, RG45 7DE, UK. Email: [email protected]. Tel/Fax: +44 (0)1344 779945.

PERMISSIONS For information on how to request permissions to reproduce articles/information from this journal, please visit www.oxfordjournals.org/permissions.

DISCLAIMER Statements of fact and opinion in the articles in Applied Linguistics are those of the respective authors and contributors and not of Applied Linguistics or Oxford University Press. Neither Oxford University Press nor Applied Linguistics make any representation, express or implied, in respect of the accuracy of the material in this journal and cannot accept any legal responsibility or liability for any errors or omissions that may be made. The reader should make his/her own evaluation as to the appropriateness or otherwise of any experimental technique described.

APPLIN-31(2)Cover.qxd

3/30/10

5:54 PM

Page 2

EDITORS

NOTES TO CONTRIBUTORS

Ken Hyland, Director, Centre for Applied English Studies, KK Leung Building, The University of Hong Kong, Pokfulam Road, Hong Kong Jane Zuengler, Nancy C. Hoefs Professor of English, University of Wisconsin-Madison, 6103 Helen C. White 600 North Park Street Madison, WI, 53706 USA Assistant to Jane Zuengler: Heather Carroll, University of Wisconsin-Madison

Articles submitted to Applied Linguistics should represent outstanding scholarship and make original contributions to the field. The Editors will assume that an article submitted for their consideration has not previously been published and is not being considered for publication elsewhere, either in the submitted form or in a modified version. Articles must be written in English and not include libelous or defamatory material. Manuscripts accepted for publication must not exceed 8,000 words including all material for publication in the print version of the article, except for the abstract, which should be no longer than 175 words. Additional material can be made available in the online version of the article. Such additions will be indexed in the print copy.

REVIEWS AND FORUM EDITOR Stef Slembrouck, Professor of English Linguistics and Discourse Analysis, Universiteit Gent, Vakgroep Engels, Rozier 44, B-9000 Gent, Belgium. <[email protected]> Assistant to Stef Slembrouck: Tine Defour, Universiteit Gent

ADVISORY BOARD Guy Cook, British Association for Applied Linguistics Aneta Pavlenko, American Association for Applied Linguistics Martin Bygate, International Association for Applied Linguistics Huw Price, Oxford University Press

EDITORIAL PANEL Karin Aronsson, Linko¨ping University David Block, London University Institute of Education Jan Blommaert, University of Jyva¨skyla¨ Deborah Cameron, University of Oxford Lynne Cameron, Open University (BAAL Representative) Tracey Derwing, University of Alberta Zolta´n Do¨rnyei, University of Nottingham Patricia Duff, University of British Columbia Diana Eades, University of New England, Australia ZhaoHong Han, Columbia University (AAAL representative) Gabriele Kasper, University of Hawai’i at Manoa Claire Kramsch, University of California at Berkeley Angel Lin, City University Hong Kong Janet Maybin, Open University, UK Tim McNamara, University of Melbourne Junko Mori, University of Wisconsin-Madison Greg Myers, Lancaster University Susanne Niemeier, University Koblenz-Landau (AILA Representative) Lourdes Ortega, University of Hawai’i at Manoa Alastair Pennycook, University of Technology, Sydney Ben Rampton, King’s College, University of London Steven Ross, Kwansei Gakuin University Alison Sealey, University of Birmingham Antonella Sorace, University of Edinburgh Lionel Wee, National University of Singapore Applied Linguistics is published five times a year in February, May, July, September and December by Oxford University Press, Oxford, UK. Annual subscription price is £254/US$457/E381. Applied Linguistics is distributed by Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. Periodicals postage paid at Rahway, NJ and at additional entry points. US Postmaster: send address changes to Applied Linguistics (ISSN 0142-6001), c/o Mercury International, 365 Blair Road, Avenel, NJ 07001, USA. # Oxford University Press 2010 All rights reserved; no part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without prior written permission of the Publishers, or a licence permitting restricted copying issued in the UK by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1P 9HE, or in the USA by the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. Typeset by Glyph International, Bangalore, India Printed by Bell and Bain Ltd, Glasgow

Applied Linguistics operates a double-blind peer review process. To facilitate this process, authors are requested to ensure that all submissions, whether first or revised versions, are anonymized. Authors’ names and institutional affiliations should appear only on a detachable cover sheet. Submitted manuscripts will not normally be returned. Forum pieces are usually reviewed by the journal Editors and are not sent for external review. Items for the Forum section are normally 2,000 words long. Contributions to the Forum section and offers to review book publications should be addressed to the Forum and Reviews Editor. For more detailed guidelines, see our website http://www.oxfordjournals.org/applij/for_authors/index.html

PROOFS Proofs will be sent to the author for correction, and should be returned to Oxford University Press by the deadline given.

OFFPRINTS On publication of the relevant issue, if a completed offprint form has been received stating gratis offprints are requested, 25 offprints of an article, forum piece or book review will be sent to the authors free of charge. Orders from the UK will be subject to a 17.5% VAT charge. For orders from elsewhere in the EU you or your institution should account for VAT by way of a reverse charge. Please provide us with your or your institution’s VAT number.

COPYRIGHT Acceptance of an author’s copyright material is on the understanding that it has been assigned to the Oxford University Press subject to the following conditions. Authors are free to use their articles in subsequent publications written or edited by themselves, provided that acknowledgement is made of Applied Linguistics as the place of original publication. Except for brief extracts the Oxford University Press will not give permission to a third party to reproduce material from an article unless two months have elapsed without response from the authors after the relevant application has been made to them. It is the responsibility of the author to obtain permission to reproduce extracts, figures, or tables from other works.

APPLIN-31(2)Cover.qxd

3/30/10

5:54 PM

Page 1

APPLI ED LI NGUISTICS

ISSN 0142-6001 (PRINT) ISSN 1477-450X (ONLINE)

Applied Linguistics

Volume 31 Number 2 May 2010

Volume 31 Number 2 May 2010

Published in cooperation with AAAL American Association for Applied Linguistics AILA International Association of Applied Linguistics BAAL British Association for Applied Linguistics

OXFORD

www.applij.oxfordjournals.org

APPLIED LINGUISTICS Volume 31 Number 2 May 2010 CONTENTS Articles Sex/Gender, Language and the New Biologism DEBORAH CAMERON The Contribution of Written Corrective Feedback to Language Development: A Ten Month Investigation JOHN BITCHENER and UTE KNOCH Probabilities and Surprises: A Realist Approach to Identifying Linguistic and Social Patterns, with Reference to an Oral History Corpus ALISON SEALEY Lexical Diversity in Writing and Speaking Task Performances GUOXING YU Discourse Particles in Corpus Data and Textbooks: The Case of Well PHOENIX W. Y. LAM An Extended Positioning Analysis of a Pre-Service Teacher’s Better Life Small Story GARY BARKHUIZEN

173

193

215 236 260

282

FORUM Input Quality Matters: Some Comments on Input Type and Age-Effects in Adult SLA JASON ROTHMAN and PEDRO GUIJARRO-FUENTES

301

REVIEWS Hannes Kniffka: Working in Language and Law: a German perspective HELEN KELLY-HOLMES

307

Nancy H. Hornberger (ed.): Can Schools Save Indigenous Languages? Policy and Practice on Four Continents JULIA SALLABANK E. Alco´n Soler and Alicia Martı´nez Flor (eds): Investigating Pragmatics in Foreign Language Learning, Teaching and Testing GILA A. SCHAUER

309

313

Jeff Siegel: The Emergence of Pidgin and Creole Languages ANDREI A. AVRAM

316

NOTES ON CONTRIBUTORS

320

Readers During 2008/2009

323

Applied Linguistics: 31/2: 173–192 ß Oxford University Press 2009 doi:10.1093/applin/amp022 Advance Access published on 31 May 2009

Sex/Gender, Language and the New Biologism DEBORAH CAMERON Oxford University, UK

INTRODUCTION This article considers some ideas about language and sex/gender,1 which are currently influential in both expert and popular discourse. Advocates of what I am calling ‘the new biologism’ (the term is discussed further below) contest the belief that male–female behavioural differences are effects of social and cultural processes. They argue instead that many such differences are biologically based, produced by evolutionary processes which have led the two sexes to differ not only in the obvious physical ways, but also in their cognitive abilities, their psychological dispositions, and consequently their habitual ways of behaving. This is not just an argument about language, but language features prominently among the phenomena to which it is applied. Here I will consider the new biologism’s claims by way of a critical examination of the arguments and the evidence. The recent resurgence of biologism in debates on sex/gender is one manifestation of a larger ‘Darwinian turn’ in the study of human behaviour, thought, and culture. The study of language and gender has not been directly affected by this development: though theoretical overviews of the field (e.g. Bing and Bergvall 1996; Bergvall 1999; McElhinny 2003) often do discuss the issue of biological sex, the point is usually to distinguish sex from gender, and to reaffirm the primacy of the latter. This research community, in short, remains committed to socio-cultural approaches.2 But in the larger academic/ scientific community, the new biologism is gaining ground. High-profile commentators like Steven Pinker (2002) cite new discoveries in rapidly advancing fields such as genetics and neuroscience to argue that biologism is now the ‘cutting edge’ approach to sex/gender, while socio-cultural approaches are

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

In recent years there has been a striking shift in both academic and popular discourse on the subject of male–female differences. It is increasingly common for biological explanations to be proposed for differences that had previously been treated by most investigators as effects of socio-cultural factors. This article critically examines the arguments as they apply to the specific case of male– female differences in linguistic behaviour. It concludes that the relevant linguistic research evidence does not on balance support the new biologism; that evidence is more adequately accounted for using the socio-cultural approaches which most linguistic researchers favour.

174 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

outmoded and discredited, maintained more for reasons of ‘political correctness’ than because the evidence supports them. Applied linguists do not dispute the general principle that scientists must be willing to follow the evidence. But in the spirit of science we should also be willing to ask what the evidence actually is, and whether it really leads where it is said to lead. In the case of language and sex/gender, how well do the new biologism’s substantive claims fit with the findings of linguistic research? How compelling is the argument that those findings are more convincingly explained in biological than socio-cultural terms? For applied linguists with their focus on language in the ‘real world’ there are reasons to broach these questions that go beyond the purely academic. The new biologism’s influence is not felt only in the academy: its arguments about sex/gender now dominate popular folklinguistic discourse, and are increasingly being taken up in the spheres of policy and professional practice. In the US, for instance, claims about innate differences between boys and girls are currently fuelling a campaign for single-sex classrooms, the idea being not just to teach girls and boys separately, but to educate them differently, with curriculum content and teaching methods tailored to suit what are alleged to be their distinctive intellectual capacities and learning styles (Gurian et al. 2001; Rivers and Barnett 2007). Supporters of this move make much of the idea that girls are innately endowed with superior verbal abilities. Boys, they say, are disadvantaged by the emphasis modern schooling places on language and literacy; they need both different language pedagogy and a less languagecentred curriculum. Elsewhere, though these ideas may not be driving educational policy, they are influencing attitudes in ways that have the potential to affect outcomes. One recent Australian study (Carr and Pauwels 2006) found both teachers and pupils invoking biological sex-differences and the notion of ‘brain sex’ to explain why girls were good at language subjects and boys were not. If we accept that educational attainment can be affected by societal expectations, and particularly by the expectations of teachers, then we might well have concerns about the current popularity of what amounts to a biologically based deficit model of boys’ verbal abilities. That model is no less in need of critical interrogation than the other verbal deficit claims (for instance, about working-class non-standard dialects, Black Englishes and ‘semilingualism’) to which linguists have turned their attention in the past. By now, readers may well be asking precisely what—or who—I mean by ‘the new biologism’. I should acknowledge that this is not a label which some particular group of scholars uses to identify itself. Rather it is an umbrella-term which I have chosen to apply to an interdisciplinary and somewhat heterogeneous collection of current scholarly enterprises, on the basis that they share certain fundamental assumptions and preoccupations, and conduct their arguments within certain parameters. It is those assumptions, preoccupations and parameters which I take to define ‘the new biologism’ as what some theorists might call a ‘discursive field’. Within that field, an important

D. CAMERON

175

shared assumption is that the most powerful explanations of the way humans think, feel, and act are those which appeal to the principles of evolutionary theory. Probably the most important source for explanatory stories of this kind—and as such, an important reference-point for the new biologism—is evolutionary psychology (hereafter ‘EP’). While few of the scholars whose work I discuss below would describe themselves as evolutionary psychologists, all of them make use of EP’s Darwinian narrative. At this point, then, I will turn to EP, setting out some of its principles and going on to consider how these are applied to the subject of language and male–female differences.

Evolutionary psychology has been described as a new ‘science of human nature’. Its central claim is that the human mind, like the human body, is a product of Darwinian evolution. On that basis it contests what its adherents represent as the hegemonic position among social scientists and laypeople, that the mind at birth is a ‘blank slate’, waiting to be inscribed by experience and socialization. Rather what humans have is an ‘adapted mind’ (Barkow et al. 1992), predisposed to develop in particular ways that proved advantageous for survival among our early human ancestors. The word ‘predisposed’ is used advisedly here: popular writers may use terms like ‘(hard) wired’ and ‘(genetically) programmed’, but most reputable scholars disclaim the strong determinism that language implies (e.g. Dunbar et al. 2007: 6–7). While such disclaimers are not unproblematic (Rose 2005), EP should not be crudely characterized as an approach that attributes everything to ‘nature’ (in the form of genes) while giving no weight to ‘nurture’ (environmental/cultural influences). On the contrary, it seeks to move beyond that dichotomy by arguing that the adaptations which produced human nature resulted from the demands made on humans by their physical and social environment interacting with biological mechanisms. The biological mechanism which is central to this story is natural selection, the process whereby genetic traits which confer a survival advantage (‘survival’ in modern evolutionary theory meaning the survival of an organism’s genes when they are passed on to offspring) gradually spread through the population. But what traits are advantageous depends on the conditions in which the organism must survive. For EP, the conditions which are relevant for our understanding of human nature are not the ones most humans live in now, but the ones in which our species evolved many millennia ago. We modern humans inherit our genes from ancestral peoples who roamed the plains of Africa in small bands of hunter-gatherers: our nature is therefore shaped by the requirements of that way of life. This general reasoning underpins the argument that cognitive and psychological differences between men and women (including the specifically linguistic ones which I will consider in more detail below) are not cultural

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

EVOLUTIONARY PSYCHOLOGY

176 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

epiphenomena, but part of our genetic inheritance. It is assumed that among early humans male and female roles were sharply differentiated: women gathered and men hunted, women nurtured infants and men fought other men. This made it adaptive for the two sexes to have different skills and traits (e.g. for women to be more empathetic and men more aggressive). In addition, the sexes are and always will be differentiated by their reproductive biology. For men, whose essential contribution to the reproductive process is fertilizing the ovum, the number of offspring who carry their genes is closely related to the frequency with which they mate. They are therefore advantaged by traits which enable them to maximize their mating opportunities by either persuading women to choose them over other men (e.g. because they are ‘better providers’) or more directly overcoming their rivals. Women, on the other hand, can only produce a limited number of offspring: for them what matters most is ensuring those they do produce survive. They are therefore advantaged by traits like empathy and nurturance which enable them to select mates judiciously and care effectively for dependent children. EP’s evocation of the lives of early humans is necessarily extrapolated from very limited direct evidence. On some questions (e.g. about our ancestors’ physiques and their diet, their tools and their visual art), conclusions may be drawn from the evidence of fossil remains and preserved artifacts. There are other questions, however, on which that evidence sheds no light, because they concern forms of behaviour which do not leave material traces. Questions about our pre-literate ancestors’ ways of using language are clearly in this category, as are questions about sexual behaviour and parenting practices. The stories EP constructs about these matters must therefore be based on more indirect evidence. The approach for which EP is best known involves identifying some trait or behaviour-pattern which is widely attested in modern human populations: on the assumption that its current prevalence (especially if it is prevalent in otherwise unrelated cultures) is most parsimoniously explained by assuming it is inherited from ancestral humans, the investigator then sets out to construct an explanation of what made it adaptive for our ancestors. It is this way of proceeding which has led critics to charge EP with telling ‘Just So Stories’. However, relevant evidence may also come from genetics, from the study of the brain and of human developmental processes, from research on nonhuman primates and from anthropological work with modern hunter-gatherer peoples, who are held to provide the closest directly observable parallel to ancestral human groups. In practice, discussions of language and sex/gender difference tend to rely on two main sources of evidence. One is the laboratory-based work of neuro- and psycholinguists investigating male–female differences in verbal ability and/or the functional organization of language in the brain. The other is research— often conducted originally by socio- or applied linguists and linguistic anthropologists—on naturally occurring verbal behaviour among men and women in

D. CAMERON

177

modern speech communities. Together, these bodies of work are taken to yield two large-scale generalizations: (a) That one sex (in most versions of the argument females, but in some versions males) is innately endowed with superior verbal abilities and a greater predisposition towards verbal communication. (b) That the two sexes differ in their typical modes of verbal interaction: men favour more competitive speech styles and genres, while women are more co-operative, empathetic, and nurturant.

SEX/GENDER AND THE EVOLUTION OF LANGUAGE There is a long history of scholarly speculation on the origins of language, and questions of sex/gender have often featured in these discussions. For instance, a once-popular story about what made the possession of a language faculty sufficiently valuable to humans to offset the cost of the large brain needed to accommodate it suggested that it enabled men to co-ordinate joint activities such as hunting and warfare more efficiently. But although there are still some theorists (e.g. Crow 1998, 2005; Miller 2000) who believe (for other reasons) that language originated in men, others now prefer accounts which suggest that it developed first in women (Knight 1990; Dunbar 1996; Joseph 2000). This preference often goes along with the view that the main adaptive value of language related to the important function (given that humans are social animals) of creating and maintaining group cohesion. Robin Dunbar (1996) argues that linguistic exchange (more specifically, the exchange of social information about community members’ activities, relationships, status, and moral worth) served the same purposes grooming serves for non-human primates. As changing environmental conditions made it desirable for the size of early human groups to increase, one-to-one physical grooming became unsustainably time-consuming. Language permitted social information to be exchanged, and relationships negotiated, in larger groups and

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

These modern generalizations are adduced as supporting evidence for the larger theory that the most reproductively successful females among our ancestors had highly developed social (including communication) skills, an innate disposition to nurture and the ability to co-operate with others, while the most successful males were effective competitors who relied more on practical and non-verbal skills. At the same time, that theory is said to offer the deepest and most powerful explanation for the modern findings. In the following discussion, I will suggest that both the generalizations themselves and the conclusions drawn from them are questionable: they are based on a very selective reading of the evidence, and in some instances also on linguistically naı¨ve or tendentious interpretations of it. I will begin, though, by considering two accounts of language and sex/gender differences which have been proposed by evolutionary scholars, and placing these in the larger context of ongoing debates about the evolution of language itself.

178 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

while engaged in other tasks. In Dunbar’s view, this account favours the hypothesis that women rather than men were the prime movers. Among non-human primates, social networks are generally centred on females. Dunbar also believes, as do other theorists who place similar emphasis on social networking/group cohesion, that his account gains support from the superior verbal skills and greater disposition to communicate which have frequently been attributed to modern human females (Halpern 1992; Kimura 1999). Scholars who take this view of language evolution have offered various proposals about how female verbal superiority evolved. One focuses on the sexual division of labour that is assumed to have prevailed among our ancestors. Joseph (2000) suggests that the habitual engagement of early human females in child-rearing, gathering, and the fashioning of domestic objects contributed to the functional evolution of the language areas of the brain, whereas men’s engagement in hunting blocked that development (gatherers and mothers can, as Joseph puts it, ‘chatter’ while they work or care for infants, but hunters must remain silent if they are not to scare off their prey). A related suggestion is that hunting developed men’s visual–spatial skills, one side-effect of which was to leave them with less neural capacity to devote to linguistic functions. But there are problems with this line of argument. Apart from the fact that modern anthropological research with hunter-gatherer peoples (Dahlberg 1981; Lee and Daly 1999) suggests that ‘men hunt, women gather’ is a considerable oversimplification, the argument relies heavily on the notion that modern women are more talkative than men; and despite being common folk-wisdom, this idea is not supported by any reputable evidence. Systematic studies using a variety of methods and measures overwhelmingly contradict it, showing that in informal peer-interaction there is typically no sex/gender difference, while in formal and status-marked situations it is most commonly men who talk more (James and Drakich 1993; Mehl et al. 2007). A similar problem arises with the claim that women favour co-operative and emotionally nurturant styles of speech whereas men are competitive and status-oriented. Unlike the female loquacity thesis, this generalization has had some support from linguistic scholarship: it was championed by the ‘difference’ current of language and gender research which was influential (though never uncontested) during the 1980s (Maltz and Borker 1982; Sheldon 1990; Tannen 1990). Difference researchers proposed that adult linguistic gender-norms reflect habits acquired in childhood, when girls and boys socialize mainly with peers of their own sex. Boys form large, hierarchical groups with leaders and followers; girls’ groups are smaller, with a looser and more egalitarian structure. Boys play competitive games with complex rules; girls play simpler games where everyone gets a turn. These differing group structures and activities teach group members to use language in different ways. Boys acquire a competitive, status-oriented communication style: they learn to argue, boast, criticize, give, and receive orders. Meanwhile, girls

D. CAMERON

179

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

acquire a more co-operative and supportive style: they learn to agree, praise, empathize, make suggestions and resolve disputes. These observations are not incompatible with an explanation in terms of innate characteristics, though that inference was not made by the original researchers. The problem, however, is that more recent research has cast doubt on the empirical validity of the difference generalizations, particularly about girls. Marjorie Harness Goodwin (2006) spent three years regularly videotaping interactions among a group of pre-adolescent girls in Los Angeles and observed that they did all the things girls had been said by difference researchers not to do. Their group had an internal hierarchy (with the lowest-status participant, a peripheral individual dubbed the ‘tag-along girl’, subjected to regular bullying by the others). When they played hopscotch or jump-rope, their goal was to win or to outshine their peers, and they frequently argued about the rules. They also engaged in regular boasting—about their skills, their possessions and the relative wealth of their families—and in issuing orders, to one another and to boys. Though Goodwin’s study stands out for the quantity and quality of the data it is based on, it is not the only case in which researchers have observed pre-adolescent and adolescent girls behaving in these ways (Eckert 1996; Baxter 2006). So far I have concentrated on the view that language evolved primarily to facilitate the (co-operative and putatively female-centred) activity of social networking. There is, however, an alternative account which emphasizes the advantages of language as a tool for courtship—a more competitive activity in which it is assumed that males took the lead. A key difference between the two accounts is that while the first centres on the mechanism of natural selection (where selected traits are those which make their possessors better adapted to survive in the prevailing environmental conditions) the second allots a central role to sexual selection, the concept which explains why traits may be selected despite conferring no obvious survival advantage. Anomalies such as the peacock’s tail (which appears to militate against survival by making the bird more conspicuous to predators) can be explained by positing that they made individuals who possessed them either more attractive to the other sex, or more successful in competing with their own sex for access to the other. This enabled them to mate more often, and so produce more offspring with the same characteristic. Sexually selected characteristics (e.g. large horns or elaborate plumage) are often found in males, and this reflects the fact that in many species it is males who do the courting while the role of females is to choose among potential mates. Peacocks, for instance, engage in ‘lekking’—ritually displaying themselves in areas frequented by peahens. Some scholars think that language fulfils analogous functions among humans. Geoffrey Miller (1999, 2000), for instance, argues that human languages are much more elaborate than they need to be to serve purely communicative purposes. This can be explained by hypothesizing that speaking served the purpose of displaying the (male) speaker’s reproductive fitness. Dunbar (1996), who believes that language evolved primarily to facilitate

180 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

EXPLAINING SEX/GENDER VARIATION: THE LIMITATIONS OF BIOLOGISM A basic assumption made by all parties to the debate is that the phenomena they discuss—instances of sex/gender-related variation in linguistic behaviour—can in principle be explained in terms of inherited biological traits. Yet as Derek Bickerton (2006) points out in a response to Locke and Bogin, this entails conflating what are arguably two different things: language itself and the uses to which it may be put. Few linguists dispute that there is a biological basis for the mental faculty which enables all developmentally normal humans to produce grammatical speech, but many would join Bickerton in questioning whether such applications of that faculty as gossiping

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

social networking, agrees that it may also have developed a secondary function as a means for men to advertise themselves to women. There is an obvious tension between this account and the orthodox wisdom which holds that women are the more talkative and verbally skilful sex. If the courtship story is correct, then men would logically have had more to gain than women from traits like loquacity, articulacy and fluency. Furthermore, on the principle that modern humans inherit the genes of the most successful reproducers among their ancestors, this male verbal advantage should still be observable. Supporters of the sexual selection account must therefore deal with the objection that the evidence points in the other direction. That objection is addressed in an article by John Locke and Barry Bogin (2006), who raise doubts about the strength of the modern psycholinguistic evidence for female verbal superiority, pointing out that meta-analytic studies have found the overall effect of sex/gender to be slight (Hyde and Linn 1988; Hyde 2005). More radically, though, these scholars question whether modern psycholinguistic findings are relevant. What bearing, they ask, can studies measuring performance on structured tests (usually requiring subjects to respond to a written stimulus under classroom-like conditions) have on debates about the innate abilities of our unschooled and pre-literate ancestors? More pertinent, they suggest, are the findings of ethnographic research in nonliterate and vernacular cultures, especially regarding oral public performance (a human analogue of lekking). These findings lead to quite different conclusions about which sex has more advanced linguistic abilities: ‘anthropological research . . . reveals that performative applications of language in the form of speech and voice consistently favor males’ (Locke and Bogin 2006: 270, emphasis in original). This internal debate among evolutionary scientists is revealing in two ways. Most obviously, it suggests a lack of consensus on some fairly fundamental issues. Yet it is also striking how much agreement there is on some of the underlying assumptions which frame the debate—assumptions which, from a linguist’s perspective, are far from self-evident.

D. CAMERON

181

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

or telling stories are themselves part of our genetic endowment. If they are not, though, then it is surely a category mistake to propose an evolutionary explanation for them. Why do scholars apparently assume that findings about sex/gender differences in the use of language are in principle susceptible to this kind of explanation? Are they suggesting there may be a gene for storytelling or a part of the brain dedicated to gossip? In most cases, no; but there is, perhaps, a clue to their reasoning in the argument about which sex is endowed with superior verbal skills. The significance accorded to this issue by both camps, even though their positions are opposed, suggests they share the assumption that observed differences in men’s and women’s linguistic behaviour must be expressions of underlying differences in the two groups’ verbal abilities. This might enable them to answer Bickerton by saying that what their account directly explains is not the behavioural differences themselves, but the innate cognitive difference which is assumed to give rise to them. But there are pressing objections to the idea that surface variation in the use of language must arise from ‘deeper’ differences in verbal ability. Students of linguistic variation do not deny that some individual language-users may be more skilful than others, but they do not link larger-scale sociolinguistic patterns to the relative verbal abilities of different groups within the population: distributional analysis suggests that the important links are to social factors such as speakers’ relative power and status, their differential access to resources like education and jobs, and their varying involvement in local networks and communities of practice (Eckert and McConnell-Ginet 1992; Holmes and Meyerhoff 1999). Evolutionary discussions of sex/gender-linked sociolinguistic patterns persistently gloss over these factors, preferring to explain variation in terms of innate cognitive differences; but their arguments are often flawed by their dependence on linguistically dubious premises. A case in point is the way much of the literature on language evolution and cognitive sex-differences treats the classical variationist generalization (now superseded, but I will leave that aside until later) that women’s speech is typically closer to the standard or prestige norm than men’s. This finding is repeatedly interpreted as evidence of women’s more advanced verbal skills. For instance, in a paragraph listing various measures on which women have been found to perform better than men in tests, Simon Baron-Cohen (2003: 60) includes the information that ‘[women’s] utterances show standard grammatical structure and correct pronunciation more often’. Doreen Kimura (1999: 91) notes that ‘girls speak more grammatically (than boys of the same age)’: later she adds that this difference is also observed in adults, which confirms that ‘speak(ing) more grammatically’ must be a reference to prescriptively defined ‘correctness’ rather than, say, the frequency of developmental errors. Evidently these scholars take the production of standard or ‘correct’ rather than non-standard or ‘incorrect’ variants as an indicator of verbal ability. For linguists, however, that is an obvious non sequitur, since judgements of correctness are based on social rather than linguistic

182 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

criteria. There is no reason to suppose that the speaker who says ‘isn’t’ or ‘haven’t’ is showing a higher level of verbal ability than the one who says ‘aint’. The problem is compounded by writers’ apparent unawareness that the sociolinguistic pattern they are discussing is not exclusively linked to sex/ gender: it also appears when researchers compare speakers from different socioeconomic classes and (in some contexts) ethnic groups. If this variation arose from innate differences in verbal ability, then presumably it would be logical to conclude that such differences exist not only between men and women, but also between middle-class and working-class speakers or white and non-white ones. It is not only linguists who would find that conclusion unpalatable. Most new biologists emphasize that they do not share the preoccupations with race and class which brought some of their predecessors into disrepute. Such obsessions, they say, were scientifically as well as politically dubious: they rested on a failure to appreciate that differences which are socially highly salient, like those between racial/ethnic groups, may nevertheless be, in Steven Pinker’s phrase, ‘biologically minor and haphazard’ (2002: 144). Male–female differences by contrast are ‘major and systematic’, making sex/gender a special case where biological explanations are legitimate. But even Pinker might find it implausible to suggest that the same linguistic pattern has a biological explanation when correlated with sex, but a cultural explanation when correlated with other demographic variables. Above I noted that the ‘women’s speech is more standard’ generalization no longer accurately represents the consensus view among variationist linguists. After reviewing the accumulated evidence, Labov (1990) formulated a set of principles drawing attention to what he has dubbed the ‘gender paradox’: in many speech communities the behaviour of women shows two opposing tendencies. On some variables women do adhere more closely than men to linguistic norms, but on others they adhere to those norms less closely. This has led variationists to reject accounts which explain women’s more ‘correct’ use of language as a consequence of their innately superior verbal abilities, since such accounts leave half the data unexplained. An alternative account which addresses this issue has been proposed by Jack Chambers (1995). He suggests that women’s superior verbal abilities are manifested not in more ‘correct’ speech, but in greater overall stylistic flexibility (i.e. women use a larger subset of the styles, genres and varieties which make up their community’s verbal repertoire). But while this might account for the patterns most commonly observed in modern western societies, it does not fit so well with the findings of cross-cultural and historical research. Linguistic anthropologists report that in many traditional non-western cultures it is men who command a wider range of styles, and who dominate or monopolize the most culturally prestigious ones (Sherzer 1987). Historical sociolinguists have made similar observations about pre-modern western communities (Nevalainen and Raumolin-Brunberg 2003).

D. CAMERON

183

SOCIO-CULTURAL APPROACHES TO SEX/GENDER VARIATION There is not just one undifferentiated ‘socio-cultural approach’ to the study of language and gender, but here I am less interested in the theoretical differences that exist within the socio-cultural camp (Note 2) than in the principles which are generally accepted by contemporary language and gender scholars. These go beyond the simple preference for socio-cultural over biological explanations of male–female differences. Most researchers also agree in rejecting the ‘essentialist’ assumption that there are characteristics (whether biologically based or socially produced) which all men or all women axiomatically share. While some take anti-essentialism to more extreme lengths than others, there has been a general retreat from the goal of making global generalizations about the linguistic behaviour of men and women. That was often the goal of early feminist research on language (though those researchers also favoured sociocultural rather than biological approaches), but today even scholars with no strong theoretical objection to it must recognize that it is in tension with the now copious empirical evidence showing that sex/gender-linked sociolinguistic patterns are variable in time and space. Accordingly, most researchers now eschew the old quest for universalizing statements, and follow instead Eckert and McConnell-Ginet’s (1992) injunction to ‘look locally’—to examine the relationship between gender and language-use in specific communities and contexts. The assumption is that similarities and differences between men and women can be related to the particularities of local social arrangements (for instance, people’s occupations, social networks, power relations, levels of literacy, rates of exogamy, beliefs about identity, etc.), but the argument is not that those arrangements determine speakers’ linguistic behaviour: rather they present men and women with

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

These observations might lead us to question not merely the arguments for female verbal superiority, but whether any account centred on innate cognitive traits can explain what needs to be explained about sex/gender-linked sociolinguistic patterns, given the empirical evidence that those patterns vary both across cultures and within them, and that they can change significantly over time. Accounts in which the behaviour of men and women arises from traits which have allegedly been part of their make-up since prehistory do not explain why the same array of male and female traits should produce such different behaviour-patterns in different times and places. Below I will expand on this argument, that the accounts proposed by the new biologism do not offer the best explanation for the empirical evidence concerning sex/gender variation, by comparing them with alternative accounts proposed by researchers who adopt a socio-cultural approach. I will begin by outlining some of the general principles which inform that approach among contemporary language and gender scholars.

184 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

particular constraints and opportunities which are part of the context for their linguistic behaviour. To illustrate how this approach works in practice, let us return to the issue of why women are found to be the more advanced users of prestigious language forms in some cases, but not in others. This variation, as already noted, presents difficulties for the argument that the use of prestige forms is connected to verbal ability. It can, however, be more convincingly related to the access women do or do not have, depending on social conditions, to the educational and professional institutions in which prestige forms are generally acquired. For example, the historical sociolinguists Terttu Nevalainen and Helena Raumolin-Brunberg (2003) found in a study of early English personal correspondence that men were more advanced than women in the adoption of high-status grammatical variants—the opposite of what would be expected now. As they point out, though, women in Tudor and Stuart England did not have the same opportunities as their present-day counterparts to acquire prestigious forms of language: even those of high social rank had very limited access to education, and none at all to the learned professions. Though in the west these restrictions have now disappeared, there are other societies where they continue to apply, and where consequently it remains the case that superordinate languages and varieties are preferentially associated with men (see e.g. Sadiqi 2003 on the case of Morocco). Even in societies where access to linguistic resources is not restricted by sex/ gender, other factors may still produce differences in men’s and women’s uptake of those resources. This point is relevant to the concern that currently exists in many western countries about the ‘gender gap’ in educational achievement—a tendency for boys’ attainment levels to lag behind those of girls, which is particularly marked in relation to language and literacy skills. Though this is often attributed to sex-differences in verbal ability, linguistic research suggests that other factors may be more significant. Some of these have to do with the salience of gender as a dimension of personal identity among children and adolescents: for many boys there is a conflict between the kind of behaviour that promotes success in school and the kind that is judged by peers to be appropriately masculine (Carr and Pauwels 2006). In addition, though, sociolinguistic patterns can often be related to the work people do, or in the case of school pupils, the work they expect to do in future; and in many cases, gender-segregated labour markets produce differing opportunities and aspirations for the two sexes. For instance, many of the ‘pink collar’ jobs available to non-elite women (e.g. in the customer service, clerical and secretarial sectors) require more standard language competence and higher levels of literacy than the ‘blue collar’ jobs open to men from the same social stratum: the latter may be able to earn as much or more in sectors like construction, which do not place a premium on language skills. Where these conditions apply, boys may invest less than girls in developing their language and literacy skills because they calculate that their prospects are less dependent on that investment. Conversely, in communities and social strata where the same conditions

D. CAMERON

185

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

do not apply, the behaviour of boys and girls will reflect that. In societies like the UK, the male underachievement pattern is far less marked among socially elite pupils—presumably because these pupils aspire to enter professions where advanced qualifications and high levels of literacy are indispensable for both sexes. These socially based accounts of sex/gender variation are, I would argue, preferable to the competing biological explanations on two main grounds. First, they are better able to deal with the empirical evidence showing that the relevant sociolinguistic patterns are not uniform cross-culturally and historically. An account which relates women’s more advanced use of prestige language forms to their innate verbal abilities has nothing to say about cases like Morocco where prestige forms are associated with men. By contrast, an account which relates prestige usage to speakers’ access to/uptake of educational and professional opportunities, and then links the sex/gender patterns observed in particular societies to the way sex/gender affects access and uptake in those societies, offers a single explanation for all the relevant empirical observations. Second, the socio-cultural approach offers a less reductive kind of explanation. The biological account by implication treats all kinds of linguistic behaviour as the natural expression of cognitive traits embodied (or ‘embrained’) in individuals; the socio-cultural one treats behaviour as the outcome of calculations and choices which, though ultimately made by individual language-users, arise within and are affected by a larger social context. The new biologists might object here that in fact their accounts give a central role to individual agency and choice—most notably mate-choice, which is treated by Darwinians as a key influence on women’s behaviour. The biological imperative to maximize reproductive success leads women to make various strategic choices with the goal of securing mates who will be ‘good providers’. In modern social conditions, one way for women to achieve that goal is through hypergamy, marriage to a man of higher social status. Robin Dunbar has argued that women’s desire to ‘marry up’ and thus advance the interests of their future children may be the underlying motivation for the sociolinguistic pattern whereby women tend to use more standard pronunciation: as he puts it, ‘It pays girls to develop an all-purpose accent that allows them to move easily up the social scale when the opportunity arises’ (1996: 185). This particular proposal is vitiated by its dependence on a gross misrepresentation of the relevant sociolinguistic pattern. Where it exists, the ‘more standard’ pronunciation of women is a matter of small differences in the frequency of particular variants (e.g. how often a /t/ sound is realized as a glottal stop); it is not the case that lower-class women speak with a non-localized accent which makes them perceptually indistinguishable from their social superordinates. However, even if Dunbar has chosen a poor illustration, the general point that women’s marital ambitions may influence their linguistic choices is one with which many linguists would concur. The question

186 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

is whether the choices women make are best explained in terms of global biological imperatives. In her classic study of language shift in an Austrian village, Susan Gal (1978) found that one key factor in the shift from Hungarian to German was the behaviour of young bilingual women whose preference for German reflected an aspiration to marry German-speaking workers rather than peasant men from their own community. However, the women’s calculations did not seem to be primarily about which men would be better providers for their future offspring. Arguably, the peasant men had advantages in this regard, since they (and later their sons) would inherit land. The women, however, were more concerned about their own quality of life: they saw marrying outside the community as a way of escaping the peasant wife’s traditional double burden of domestic and agricultural labour. Similarly self-interested motives are suggested by Ingrid Piller and Kimie Takahashi’s study (2006) of a group of Japanese women who had invested significant resources in acquiring English because of their desire to contract relationships with Anglophone men. Once again, it is not obvious that these women’s preferences reflected the ‘good provider’ principle (Japan has both higher levels of job-security for men and lower rates of divorce than Anglo-Saxon cultures). Rather they were based on a perception of Anglophone men as more glamorous than Japanese, and less traditional in their attitudes to women. In all the cases mentioned so far, it is evident that the constraints negotiated by women (e.g. restricted educational access, gender-segregated labour markets, economic dependence on marriage) are connected not merely to gender difference but also to gender hierarchy (albeit different societies exhibit this in varying degrees and forms). Socio-cultural approaches treat power relations between men and women in a given community as potentially an important influence on their linguistic behaviour. The new biologism, on the other hand, tends either to overlook the effects of male dominance or else to reinterpret them as manifestations of innate difference. Once again, it may be argued that the resulting explanations of behaviour are reductive and inadequate. A phenomenon which many researchers have related to power differences is the sex/gender-linked patterning of linguistic politeness behaviour. Evolutionary scholars, however, often refer to these researchers’ findings while discarding their explanations. Thus Dunbar’s (1996) discussion of Penelope Brown’s (1980) study of politeness behaviour in a Mayan community treats her finding that women are ‘more polite’ (i.e. use higher frequencies of certain politeness features) as yet another piece of evidence for their superior verbal abilities. In Brown’s own analysis, by contrast, the issue is not ability, it is motivation; and the motivation arises from inequality. In this community women’s subordination is extreme, and non-compliance with men’s demands carries a high risk of physical punishment. Brown suggests that this explains why women’s greater politeness is not just a global, undifferentiated phenomenon: their preference for using deferential forms of politeness to men, and more solidary forms to other women, becomes

D. CAMERON

187

SEX, LANGUAGE AND THE BRAIN Many discussions of sex, language and the brain focus on the phenomenon of lateralization, the specialization of functions to one or other of the brain’s two hemispheres. Language is (in right-handed individuals) predominantly a left-hemisphere function. But there has long been interest in the possibility that the functional organization of the brain, including its organization for language, may be different for males and females, with males (on average, and other things—like hand-preference—being equal) showing a greater degree of lateralization. Evidence bearing on this question comes from neuroimaging studies, experiments using the Wada technique, and clinical observations of brain-injured patients. A related area of investigation concerns the effect of hormones. It has been suggested that high levels of testosterone (a hormone which is found in both sexes, but normally at significantly higher concentrations in males) produce earlier and more rapid growth on the right side, including the right hemisphere

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

understandable if we view politeness as a strategy they employ to deal with problems that arise from their social position. With men it is rational to try to disarm potential threats by displaying a submissive or non-provocative attitude; with other women it is rational to try to form protective alliances by displaying solidarity and mutual regard. Men are ‘less polite’ not because they cannot use these strategies, but because in most situations they feel no need to. The same pattern would not be expected in societies where male–female relationships are organized differently: indeed, there are societies where the most elaborately polite linguistic styles are dominated or monopolized by men while verbal aggression or impoliteness is associated with women (Keenan 1974; Kulick 1993). Above I argued that the new biologism’s accounts of sex/gender differences in linguistic behaviour are flawed by their reliance on generalizations which the evidence does not support (such as the belief that women talk more than men), and/or assumptions which are linguistically ill-founded (such as the equation of ‘correctness’ with verbal skill). Here I have drawn attention to other problems, such as inattentiveness to cross-cultural and historical evidence, and failure to consider (or in some cases even to acknowledge) competing interpretations of data. The most pertinent objection to these accounts is not that they offend feminist and other ‘politically correct’ sensibilities; it is that they do not meet the scientific criterion of being able to account fully and convincingly for the evidence. But as I noted in the introduction, there is another body of evidence to which the new biologism appeals, drawn from the study of language, cognition and the brain. Findings about what popularizers have dubbed ‘brain sex’ are often invoked, in both scholarly and non-scholarly sources, as if they constituted irrefutable evidence for the reality and significance of biologically based sex-differences. How far, though, is that interpretation justified?

188 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

the widely held assumption that there are major differences in the degree of brain lateralization of function between men and women is debatable. Even where sex differences in lateralization clearly exist, no convincing case has yet been made that they influence cognitive function. In this area of research it is possible that future advances will yield firmer conclusions (whereas we would need to solve the problem of time-travel to get much further with the study of early human linguistic behaviour). But there will still be questions about how far neuroscientific findings license the broader claims they are often invoked to support. How, that is, do we get from specific statements about the performance of men and women on measures X, Y, and Z (as demonstrated in laboratory tasks designed so that investigators can isolate a particular skill, observe it systematically and measure it reliably), to much more general assertions about women’s superior communication skills when interacting with others in everyday life? Kimura is typical of many commentators on this subject in treating the connection as obvious. ‘The impression many people have’, she remarks (1999: 91), ‘that women generally possess better skill with words than men do, probably arises from differences apparent between very young girls and boys’. She then lists some differences for which there is experimental evidence: girls have larger working vocabularies, are better spellers, superior readers, and do better on ‘tests requiring them to generate words with particular limitations

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

of the brain, and that this has implications for our understanding of both normal male–female differences and the aetiology of certain developmental abnormalities (e.g. Geschwind and Galaburda 1985). Scholars who favour this argument relate it to an observation whose centrality to some evolutionary accounts has already been discussed: that in comparison to most females, most males have less well-developed verbal abilities, but better-developed visual–spatial skills. If the testosterone theory is correct it may explain this, since linguistic functions are typically concentrated in the left hemisphere and visual-spatial functions in the right one. However, its correctness remains a matter of dispute (e.g. Crow 1998), as indeed does the assertion that females have more advanced verbal skills (see above). On lateralization too, many questions remain unresolved. Recent advances in the study of the brain (especially the advent of new imaging techniques) have produced a wealth of new data, but it has not necessarily become easier to draw clear conclusions from it: studies have proliferated, but their findings are very mixed. Supporters of the testosterone theory can appeal to studies in which hemispheric dominance for language was found to be stronger in males; but sceptics can equally point to studies using the same tasks which found no sex-differences (Frost et al. 1999; Knecht et al. 2000). Even Doreen Kimura, an influential supporter of the general thesis that there are innate cognitive differences between the sexes, has urged caution when interpreting this body of evidence (1999: 181):

D. CAMERON

189

CONCLUSION The new biologism is part of a larger ‘Darwinian turn’, in the light of which linguists (among others) are now frequently urged to re-examine their assumptions about the nature and behaviour of human beings. I do not dispute the importance of paying attention to developments in other fields which may have implications for our own. But I do think there is a need for caution about the rhetorical claims which are often made by and for the new biologism. In this discussion I have focused specifically on its claims about sex/gender differences in linguistic behaviour, and I have argued that its approach to this topic is not preferable on scientific criteria to the socio-cultural approaches favoured by most current language and gender researchers. I have tried to show that the latter explain more of the relevant empirical evidence with less resort to assumptions and speculations for which there is no evidential basis. I have also drawn attention to the selective and misleading way in which linguistic research evidence is often used by supporters of the new biologism. When these scholars accuse others of being unwilling to follow the evidence where it leads (i.e. to the conclusion that male–female differences are biologically based), they are open to the charge of throwing stones from a glass house. Scientific accountability is not a one-way street: if linguists must take account of new knowledge about genes, brains and evolution, then Darwinians must equally be held accountable to the knowledge our research has produced about language-use and linguistic variation. It is right that we should learn from other scientists’ work; but that need not and should not mean devaluing our own. Conflict of interest statement. None declared.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

on them (such as words containing particular letters)’, though they are ‘not necessarily more fluent in narrative output’ (1999: 91). But on reflection it is hard to see how most of these variables could be directly involved in the everyday impressionistic judgements which Kimura suggests they may explain. Spelling and reading skills are not displayed in conversation, and most conversations do not require participants to display more than a small part of their working vocabulary, let alone ‘generate words with particular limitations on them’. The only variable mentioned which does seem plausible as a basis for real-world judgements of skill is narrative fluency, which is also the only variable on which Kimura says there is no clear evidence of female superiority. Does she take the variables which have shown sex/gender differences in the lab to be legitimate proxies for other skills that might be observable outside it, and if so what skills would those be? Just placing one set of observations (many people have the impression that. . . ) adjacent to another set (laboratory tests have shown that. . . ) does not magically make one into the explanation of the other.

190 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

NOTES 1

is influenced by postmodernist theories of identity and subjectivity, or alternatively by ethnomethodological theories of the construction of social reality. The distinctive feature of this approach is that it denies gender any ontological reality (here I paraphrase the postmodernist philosopher Judith Butler)—in other words, proposes that ‘men’ and ‘women’ are effects of our discourses and practices rather than entities which exist independently of those discourses and practices. For the most radical ‘gender relativists’, an account which invokes external social structures, hierarchies or norms to explain gendered behaviour is not ‘social constructionist’, because it requires the analyst to posit that gender or some aspect thereof does have a pre- or extra-discursive reality (this position is labelled ‘gender realism’). I am a ‘gender realist’, and I do not accept that this position is not also ‘social constructionist’ in the classical Beauvoirian sense. However, since the meaning of ‘social constructionism’ is now contested, and for some people the term does exclude non-relativist or non-postmodern approaches, I have tried to pre-empt any misunderstanding by using the term ‘socio-cultural approach’. I intend this to function simply as a generic label for the nonbiologically based approaches to sex/ gender which this article contrasts with the new biologism.

REFERENCES Barkow, J., L. Cosmides, and J. Tooby (eds). 1992. The Adapted Mind: Evolutionary Psychology

and the Generation of Culture. Oxford University Press.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

2

The slightly cumbersome formula ‘sex/ gender’ is used here and (where appropriate) below in an effort to avoid misrepresenting the competing positions in this debate, or pre-empting the argument between them. The term ‘gender’ (meaning ‘the social condition of being a man/woman’) is generally preferred to ‘sex’ in social scientific and/or feminist research, but the new biologism argues that many or most socalled ‘gender differences’ are really effects of biological sex. While it will be clear that I disagree with the new biologism, I have tried to avoid using terminology which inherently favours one position. Some readers may wonder why I refer throughout this article to ‘socio-cultural approaches’ rather than ‘social constructionism’. In my own usage, in fact, these expressions would be more or less interchangeable. Until quite recently, the description of feminist theory or scholarship as ‘social constructionist’ implied only that the theorist or scholar accepted Simone de Beauvoir’s foundational insight (first aired in 1949) that ‘one is not born a woman, one becomes one’— in other words that men, women and the differences between them are the products of society and culture, not biology. Since the 1990s, however, the term ‘social constructionism’ has undergone narrowing: some feminist scholars now use it to refer specifically and exclusively to one particular variant of social constructionism, which

D. CAMERON

Belief Systems. Berkeley Women and Language Group, pp. 183–90. Eckert, P. and S. McConnell-Ginet. 1992. ‘Think practically and look locally: Language and gender as community-based practice,’ Annual Review of Anthropology 12: 461–490. Frost, J. A., J. R. Binder, J. A. Springer, T. A. Hammeke, P. S. Bellgowan, S. M. Rao, et al. 1999. ‘Language processing is strongly left lateralized in both sexes: Evidence from functional MRI,’ Brain 122: 199–208. Gal, S. 1978. ‘Peasant men can’t get wives: language and sex roles in a bilingual community,’ Language in Society 7/1: 1–17. Geschwind, N. and A. M. Galaburda. 1985. ‘Cerebral lateralization. Biological mechanisms, associations and pathology: A hypothesis and a program for research,’ Archives of Neurology 42: 428–654. Goodwin, M. H. 2006. The Hidden Life of Girls: Games of Stance, Status and Exclusion. Blackwell. Gurian, M., P. Henley, and T. Trueman. 2001. Boys and Girls Learn Differently! Wiley. Halpern, D. F. 1992. Sex Differences in Cognitive Abilities. Lawrence Erlbaum. Holmes, J. and M. Meyerhoff. 1999. ‘The community of practice: theories and methodologies in language and gender research,’ Language in Society 28: 173–183. Hyde, J. S. 2005. ‘The gender similarities hypothesis,’ American Psychologist 60/6: 581–592. Hyde, J. S. and M. Linn. 1988. ‘Gender differences in verbal ability: A meta-analysis,’ Psychological Bulletin 104: 53–69. James, D. and J. Drakich. 1993. ‘Understanding gender differences in amount of talk’ in D. Jannen (ed.): Gender and Conversational Interaction. Oxford University Press, pp. 281–312. Joseph, R. 2000. ‘The evolution of sex differences in language, sexuality and visualspatial skills,’ Archives of Sexual Behavior 29/1: 35–66. Keenan, E. O. 1974. ‘Norm makers, norm breakers: Some uses of speech by women in a Malagasy community’ in R. Bauman and J. Sherzer (eds): Explorations in the Ethnography of Speaking. Cambridge University Press, pp. 125–43. Kimura, D. 1999. Sex and Cognition. MIT Press. Knecht, S., M. Deppe, B. Draeger, L. Bobe, H. Lohmann, E.-B. Ringelstein, and

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Baron-Cohen, S. 2003. The Essential Difference: Men, Women and the Extreme Male Brain. Allen Lane. Baxter, J. 2006. ‘ ‘‘Do we have to agree with her?’’ How high school girls negotiate leadership in public contexts’ in J. Baxter (ed.): Speaking Out: The Female Voice in Public Contexts. Palgrave Macmillan, pp. 159–78. Bergvall, V. L. 1999. ‘Toward a comprehensive theory of language and gender,’ Language in Society 28: 273–293. Bickerton, D. 2006. ‘Language-use, not language, is what develops in childhood and adolescence,’ Behavioral and Brain Sciences 29: 280–281. Bing, J. M. and V. L. Bergvall. 1996. ‘The question of questions: beyond binary thinking’ in V. L. Bergvall, J. M. Bing, and A. F. Freed (eds): Rethinking Language and Gender Research: Theory and Practice. Longman, pp. 1–30. Brown, P. 1980. ‘How and why are women more polite? Some evidence from a Mayan community’ in S. McConnell-Ginet, R. Borker, and N. Furman (eds): Women and Language in Literature and Society. Praeger, pp. 111–36. Carr, J. and A. Pauwels. 2006. Boys and Foreign Language Learning: Real Boys Don’t Do Languages. Palgrave Macmillan. Chambers, J. K. 1995. Sociolinguistic Theory. Blackwell. Crow, T. 1998. ‘Sexual selection, timing and the descent of man: A theory of the genetic origins of language,’ Cahiers de Psychologie Cognitive/Current Psychology of Cognition 17/6: 1079–1114. Crow, T. 2005. ‘Who forgot Paul Broca? The origins of language as test case for speciation theory,’ Journal of Linguistics 41: 133–156. Dahlberg, F. (ed.) 1981. Woman the Gatherer. Yale University Press. Dunbar, R. 1996. Grooming, Gossip and the Evolution of Language. Faber. Dunbar, R., L. Barrett, and J. Lycett. 2007. Evolutionary Psychology, a Beginner’s Guide: Human Behaviour, Evolution and the Mind. Oneworld. Eckert, P. 1996. ‘Vowels and nail polish: the emergence of linguistic style in the preadolescent heterosexual marketplace’ in N. Warner, J. Ahlers, L. Bilmes, M. Oliver, S. Wertheim, and M. Chen (eds): Gender and

191

192 SEX/GENDER, LANGUAGE AND THE NEW BIOLOGISM

Miller, G. F. 1999. ‘Sexual selection for cultural displays’ in R. Dunbar, C. Knight, and C. Power (eds): The Evolution of Culture. Edinburgh University Press, pp. 71–91. Miller, G. F. 2000. The Mating Mind: How Sexual Choice Shaped the Evolution of Human Nature. Doubleday. Nevalainen, T. and H. Raumolin-Brumberg. 2003. Historical Sociolinguistics. Longman. Piller, I. and K. Takahashi. 2006. ‘A passion for English: Desire and the language market’ in A. Pavlenko (ed.): Bilingual Minds: Emotional Experience, Expression and Representation. Multilingual Matters, pp. 59–83. Pinker, S. 2002. The Blank Slate: The Modern Denial of Human Nature. Allen Lane. Rivers, C. and R. C. Barnett. 2007. ‘The difference myth,’ Boston Globe, October 28. Rose, S. 2005. The 21st Century Brain: Expanding, Manipulating and Mending the Mind. Vintage. Sadiqi, F. 2003. Women, Gender and Language in Morocco. Brill. Sheldon, A. 1990. ‘Pickle fights: gendered talk in pre-school disputes,’ Discourse Processes 13: 5–31. Sherzer, J. 1987. ‘A diversity of voices: women’s and men’s speech in ethnographic perspective’ in S. Phillips, S. Steele, and C. Tanz (eds): Language, Gender and Sex in Cross-Cultural Perspective. Cambridge University Press, pp. 95–120. Tannen, D. 1990. You Just Don’t Understand: Men and Women in Conversation. Morrow.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

H. Henningsen. 2000. ‘Language lateralization in healthy right-handers,’ Brain 123: 74–81. Knight, C. 1990. Blood Relations: Menstruation and the Origins of Culture. Yale University Press. Kulick, D. 1993. ‘Speaking as a woman: structure and gender in domestic arguments in a Papua New Guinea village,’ Cultural Anthropology 8/4: 510–541. Labov, W. 1990. ‘The intersection of sex and social class in the course of linguistic change,’ Language Variation and Change 2: 205–254. Lee, R. and R. Daly (eds) 1999. The Cambridge Encyclopaedia of Hunters and Gatherers. Cambridge University Press. Locke, J. and B. Bogin. 2006. ‘Language and life-history: A new perspective on the evolution and development of human language,’ Behavioral and Brain Sciences 29: 259–280. Maltz, D. and R. Borker. 1982. ‘A cultural approach to male-female misunderstanding’ in J. J. Gumperz (ed.): Language and Social Identity. Cambridge University Press, pp. 196–216. McElhinny, B. 2003. ‘Gender in sociolinguistics and anthropology’ in J. Holmes and M. Meyerhoff (eds): The Handbook of Language and Gender. Blackwell, pp. 21–42. Mehl, M. R., S. Vazire, N. Ramı´rez-Esparza, R. B. Slatcher and J. W. Pennebaker. 2007. ‘Are women really more talkative than men?,’ Science July 6: 82.

Applied Linguistics: 31/2: 193–214 ß Oxford University Press 2009 doi:10.1093/applin/amp016 Advance Access published on 20 May 2009

The Contribution of Written Corrective Feedback to Language Development: A Ten Month Investigation 1

JOHN BITCHENER and 2UTE KNOCH

1

AUT University, Auckland, New Zealand, 2University of Melbourne, Australia

INTRODUCTION The call for longitudinal research into the efficacy of written corrective feedback (WCF) can be traced back to the debate between Truscott and Ferris in the mid- to late 1990s. In 1996, Truscott claimed that error correction in ESL (English as a second language) writing programmes should be abandoned because it is ineffective and harmful. He claimed (i) that there was no research evidence to support the view that it ever helps student writers; (ii) that, as typically practised, it overlooks SLA (second language acquisition) insights about how different aspects of language are acquired; and (iii) that practical problems related to how teachers provide WCF and how students receive it to make a futile endeavour. However, in his 1999 response to Ferris (1999), he admitted that future research should look for, and might well find, some special, narrow uses of correction which may have value. He also claimed that it is harmful because it diverts time and energy away from more productive aspects of writing instruction. However, Ferris (1999) pointed out that the claims were premature because the body of evidence he presented was too limited and because there were too many methodological flaws in the design and analysis of the published studies. She also explained that short-term investigations involving text revision reveal improvement in accuracy as a result of WCF and that students believe it helps them improve their writing.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The call for longitudinal evidence on the efficacy of written corrective feedback (WCF) for ESL (English as a second language) writers has been made repeatedly since Truscott (1996) claimed that it is ineffective, harmful, and should therefore be abandoned. This article discusses some of the theoretical issues raised against the practice, outlines the status of recent empirical evidence and presents a 10-month study of the effects of WCF on two functional uses of the English article system given to 52 low-intermediate ESL students in Auckland, New Zealand. Assigned to four groups (direct corrective feedback, written, and oral meta-linguistic explanation; direct corrective feedback and written meta-linguistic explanation; direct corrective feedback only; the control group), the students produced five pieces of writing (pre-test, immediate post-test, and three delayed post-tests). Each of the treatment groups outperformed the control group on all post-tests and no difference in effectiveness was found between the three treatment groups.

194 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

THEORETICAL ISSUES RAISED BY TRUSCOTT From a theoretical perspective, Truscott (1996) claimed that there are a number of reasons for not expecting WCF to be effective. Agreeing with well-established SLA claims about the gradual and complex nature of the acquisition process, he built his first argument against the value of WCF around the belief that a simple transfer of information (in the form of WCF) cannot be expected to work. However, this stance does not take into account the fact that learners who notice the difference between target-like input (be it oral or WCF) and their non target-like output are able to modify it as targetlike output. There is sufficient evidence in both the oral SLA literature (Schmidt and Frota 1986; Swain 2005) and the text revision literature (Ferris 2004, 2006) to confirm that this can occur. Moreover, in recent WCF studies (to be discussed in the following section), it has been reported that learners are able to apply the feedback given to them on targeted linguistic forms/structures when writing new texts over at least a 6-month period (Bitchener et al. 2005; Sheen 2007; Bitchener 2008; Bitchener and Knoch 2008a, 2008b, 2009). However, further research is required to see what enduring effect WCF might have beyond this period of time, thereby determining whether there is a role for the practice beyond text revision and short-term development.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

However, she acknowledged that some of Truscott’s reasoning was compelling, particularly with regard to some of the theoretical issues he raised and to some of the practical problems he outlined. Ferris concluded her response with a call for further research into questions about the efficacy and provision of WCF on ESL student writing. While agreeing with this agenda, Truscott (1999) insisted and continues to insist (Truscott 2004, 2007) that until clear evidence is produced that WCF is effective and it must be assumed that it is a misguided practice. To date, Truscott’s theoretical arguments against WCF have received limited response in the literature. On the other hand, the call by both Truscott and Ferris for more empirical evidence on the long-term benefits of WCF has resulted in a growing number of published studies. However, the extent to which it has an enduring effect beyond a 2-month period has yet to be explored. A range of studies have investigated the relative effects of direct and indirect feedback options but the results from these studies have varied so much that further investigations are required before conclusions can be made. To address these needs, this article (i) responds to some of the theoretical issues identified by Truscott; (ii) surveys the current status of empirical evidence (on the efficacy of WCF and on the relative effect of different feedback options); and (iii) presents the findings of a 10-month investigation into the effectiveness of providing WCF on two targeted linguistic error categories, measured by the level of accuracy achieved in the writing of new texts.

J. BITCHENER and U. KNOCH

195

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Truscott further argues that no single form of correction can be expected to help learners acquire knowledge of all linguistic forms/structures because the acquisition of syntax, morphology, and lexis require an understanding not only of form but also of meaning and use in relation to other words and other parts of the language system. It is important that future research agendas include investigations of the extent to which WCF, and different types of WCF, can facilitate the acquisition of different linguistic forms/structures. To date, there is research evidence to suggest that WCF can be helpful in the acquisition of simple, rule governed forms/structures like aspects of the English article system and the past simple tense (Bitchener et al. 2005; Bitchener 2008). Concerning the effectiveness of WCF in other linguistic areas, further research is required to see if there are limitations on its wider effectiveness. Truscott suggests that WCF cannot be expected to facilitate syntactic knowledge because the domain is not characterized by a collection of discreet items that can be learned one by one. Given the positive effects that have been observed in oral corrective feedback studies, for example, with syntactic structures like question forms (Mackey and Philp 1998; Mackey and Oliver 2002; Mackey et al. 2002) and the use of dative constructions (McDonough 2006), it is possible that further WCF research may also reveal positive effects for these and other forms/structures. Moreover, it may well be the case that WCF proves to be more effective as a result of the additional time and space ESL writers have to consider and process it. The second reason that Truscott presents against the practice of WCF concerns the feasibility of providing WCF at a time when the learner is ‘ready’ (Piennemann 1998) to acquire a particular form/structure. Because the acquisition of some linguistic forms and structures has been shown to follow a natural order (Clahsen et al. 1983), it is understandable that WCF may not be effective if it is provided at a time that is inconsistent with that order. However, if teachers take into account a learner’s current stage of development when determining their areas of focus, the potential would always exist for it to be effective. Concerning this, Truscott questions the feasibility of providing learners with WCF at a time that coincides with their readiness. However, this does not mean that the task is impossible. It is possible, for example, in both individual and small group contexts for WCF to be effectively provided on one or two targeted forms/structures that learners and teachers identify as being repeatedly problematic and that learners agree should be targeted for an agreed period of time. Truscott continues his argument by suggesting that WCF is unlikely to be effective if it is provided on too comprehensive a range of error categories. While the comprehensive approach may not prevent learners from accurately revising a number of errors in a single text, it may not enable learners to demonstrate equally positive improvements in their writing of new texts over time. On the other hand, Truscott also discounts a more selective approach (whereby a few error categories are consistently corrected over a period of time), claiming that there is evidence to suggest its ineffectiveness. However, he refers to only one study

196 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

EMPIRICAL EVIDENCE IN SUPPORT OF WCF As well as presenting theoretical reasons to suggest that WCF is unlikely to be effective, Truscott claimed that several empirical studies (Semke 1984; Robb et al. 1986; Kepner 1991; Sheppard 1992) had demonstrated that there

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(Hendrickson 1981). Before such a claim can be made, evidence from single study investigations that compare the effect of WCF on targeted and untargeted categories is needed. Just as positive findings have been reported in a number of oral corrective feedback studies where single error categories have been targeted (Doughty and Varela 1998; Han 2002; Iwashita 2003; Lyster 2004), so it may also be the case with similar WCF investigations. The third argument that Truscott presents against the practice of WCF is that any learning that results from the practice is likely to be only ‘pseudolearning’, described by Truscott (1996: 345) as ‘a superficial and possibly transient form of knowledge’. The distinction between knowledge of language and knowledge about language is one that has been made elsewhere in the literature. Krashen (1982, 1985, 1994, 1999) distinguished between acquisition and learning while DeKeyser (2004), Ellis (2004), and others have distinguished between implicit unconscious procedural knowledge and explicit declarative knowledge. Truscott argues that, at best, WCF may have some limited value for developing meta-linguistic knowledge or explicit declarative knowledge and, therefore, may be useful for editing purposes. The value of WCF for editing purposes has been empirically demonstrated (Ferris 2004, 2006) and, in the case of some error categories (see next section), shown to also have value for language development. However, the durability of these effects over longer periods of time has yet to be investigated. The extent to which learners draw upon explicit or implicit knowledge during the writing process is not known. It is likely, however, that they draw upon both explicit and implicit linguistic knowledge. At certain times during the writing process, they are likely to reflect upon the form or structure that should be used, and, in such situations, they are most likely to draw upon their explicit, meta-linguistic knowledge, whereas, at other times, the decision about which form or structure to use will be made without conscious reflection. Whether the WCF that is used to edit a single text (Ferris 2003) or write a new text over time (Sheen 2007; Bitchener 2008; Bitchener and Knoch 2008a, 2008b, 2009) is stored in the learner’s memory as explicit declarative knowledge or as implicit procedural knowledge has yet to be investigated. However, there is a growing longitudinal research base to show that ESL writers make use of the WCF that they are given when it comes to the writing of new texts over time. This demonstrates that improvements in accuracy can result from WCF. Thus, it is incorrect to suggest that, in such cases, only pseudo-learning has taken place. Improved accuracy in the writing of new texts over time demonstrates more than a superficial and transient level of knowledge. The following section outlines the status of such evidence.

J. BITCHENER and U. KNOCH

197

is evidence to support an abandonment of the practice. Given the design and analytical shortcomings of these studies (Ferris 2003; Guenette 2007; Sheen 2007; Bitchener and Knoch 2008a), such conclusions need to be read with caution. Since making this claim, an expanding research base has been considering (i) whether or not WCF can help learners improve the accuracy of their writing and thereby facilitate the acquisition process and (ii) the relative effectiveness of different types of WCF.

The effectiveness of WCF

The relative effectiveness of different types of WCF Assuming that WCF is effective in helping learners improve the accuracy of their writing and in facilitating the acquisition process, a range of studies have investigated whether certain types of WCF or combinations of different types

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

A number of studies conclude that WCF is effective in helping ESL students improve the accuracy of their writing. However, seven of these (Lalande 1982; Ferris 1995, 1997, 2006; Chandler 2000; Ferris and Helt 2000; Ferris et al. 2000) were designed without a control group so there is no way of knowing whether or not the reported improvements in accuracy were actually a result of WCF. Of those that did compare groups of students who received WCF and those who did not, eight (Fathman and Whalley 1990; Ashwell 2000; Ferris and Roberts 2001; Sheen 2007; Bitchener 2008; Bitchener and Knoch 2008a, 2008b, 2009) report that it had a positive effect on accuracy. Studies by Ashwell (2000), Fathman and Whalley (1990), and Ferris and Roberts (2001) reported improved accuracy in text revisions while those by Bitchener (2008), Bitchener and Knoch (2008a, 2008b, 2009), and Sheen (2007) report improved accuracy in the writing of new texts over time. On the other hand, two other studies (Kepner 1991; Polio et al. 1998) that included a control group reported that WCF was not effective in helping students improve the accuracy of their writing. However, it needs to be realized that both of these studies contained design flaws so each should be read with this in mind. Kepner’s study had no pre-test measurement, no control over the length of the journal entries, no control over the texts that were written out of class, and a range of analytical flaws. Less problematic was the study by Polio et al. where different instruments were included in the post-test. Other studies (Lalande 1982; Chandler 2000; Ferris 2006) that have not included a control group have also claimed that WCF facilitates improvements in accuracy but, without a non-feedback group, they can only be regarded as possibly predictive of effectiveness. Further research that includes a control group, that tests the efficacy of WCF in new pieces of writing and that measures the level of retention over more extensive periods of time is needed if conclusions about the acquisition potential of WCF, as opposed to the text revision potential, are to be made.

198 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

are more effective than others. These studies have most often categorized feedback as either direct (explicit) or indirect (implicit). Direct corrective feedback has most often been defined as the provision of the correct linguistic form or structure by the teacher to the student above or near the linguistic error. It may include the crossing out of an unnecessary word/ phrase/morpheme, the insertion of a missing word/phrase/morpheme, or the provision of the correct form or structure. More recently, direct corrective feedback has also included written meta-linguistic explanation (the provision of grammar rules and examples of correct usage). Sometimes oral metalinguistic explanation (e.g. in the form of class discussion or one-on-one conferences) has also been included in this category. On the other hand, indirect corrective feedback is that which indicates that in some way an error has been made, but correction is not supplied. This may be provided in one of four ways: underlining or circling an error; recording in the margin the number of errors in a given line; or using a code to show where an error has occurred and what type of error it is. Rather than the teacher providing an explicit correction, students are left to resolve and correct the problem that has been drawn to their attention. Studies that have investigated the relative merits of these approaches have tended to be grouped according to those that have compared (i) direct and indirect types of WCF; (ii) different types of indirect feedback; and (iii) different types of direct feedback. In studies that have compared direct and indirect types, two (Lalande 1982) have reported an advantage for indirect feedback, two (Semke 1984; Robb et al. 1986) have reported no difference between the two approaches, and one (Chandler 2003) has reported positive findings for both direct and indirect feedback. Clearly, firm conclusions cannot be made from these conflicting results. In comparison, findings from some oral corrective feedback studies in SLA research point to an advantage for direct over indirect corrective feedback (Carroll and Swain 1993; Nagata 1993; Muranoi 2000; Carroll 2001; Havranek and Cesnik 2003; Rosa and Leow 2004; Ellis et al. 2006), but there are others (Dekeyser 1993; Kim and Mathes 2001; Leeman 2003) that claim the opposite. As well as comparing direct and indirect approaches, several other studies (Robb et al. 1986; Ferris et al. 2000; Ferris and Roberts 2001) have investigated the relative effectiveness of different types of indirect feedback (coded and uncoded), but none has found any difference between the two options. Less attention has been given to a comparison of different types of direct feedback. Bitchener et al. (2005) compared the effect of different direct feedback combinations typically practised in advanced proficiency classroom settings: direct correction plus oral meta-linguistic explanation in the form of five minute oneon-one conferences; direct error correction; no corrective feedback. They found that group 1 (direct error correction and oral meta-linguistic explanation) outperformed group 2 (direct error correction) and the control group for the past simple tense and the definite article but not for prepositions.

J. BITCHENER and U. KNOCH

199

THE STUDY Aims Two aims informed the design of this study. The first was to add to the growing body of research investigating the extent to which WCF results in improved accuracy in four new pieces of writing over a 10-month period and, as a result, to determine the role that WCF might play in long-term acquisition. The second aim was to investigate whether there is a differential effect on accuracy for three different WCF options: (i) error correction plus written and oral meta-linguistic explanation; (ii) error correction plus written meta-linguistic

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

They suggest that the addition of oral meta-linguistic explanation may have been the crucial factor in facilitating increased accuracy. Bitchener (2008) investigated the effectiveness of other direct feedback combinations (direct error correction with written and oral meta-linguistic explanation in the form of a 30-minute classroom lesson; direct error correction with written meta-linguistic explanation; direct error correction; no corrective feedback) on only two functional uses of the English article system (indefinite ‘a’ for first mention and definite ‘the’ for subsequent mentions) with 75 low intermediate, international visa students. Groups 1 and 3 outperformed the control group while group 2 (direct error correction and written meta-linguistic explanation) only just failed to do so. When the study was extended (Bitchener and Knoch 2008a) to include 144 international visa and migrant students, no difference was observed between the same three treatment combinations. It is possible that the larger sample size eliminated the difference in effect between group 2 and the other two treatment groups in the first study by Bitchener (2008). Sheen (2007) investigated the relative effect of two types of direct feedback (error correction and written meta-linguistic explanation) with 91 intermediate ESL learners and found that those who received WCF outperformed those who received no feedback. She also found no difference between the two feedback options in her immediate post-test, but an advantage for written meta-linguistic explanation over direct error correction in the delayed posttest conducted 2 months later. Considering four studies, it is clear that focused WCF was more effective than no feedback in bringing about improvements in the accuracy of ESL learners. Whether or not there is an advantage for meta-linguistic explanation over error correction alone for some forms/structures has yet to be confirmed. On the one hand, Sheen (2007) found an advantage for meta-linguistic explanation but this only became evident 2 months later in her delayed post-test. On the other hand, Bitchener and Knoch (2008a, 2008b) found no advantage for those who received meta-linguistic explanation after a similar 2-month period. Further research that addresses these factors over time may enable firmer conclusions to be drawn.

200 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

explanation; and (iii) error correction. Two research questions were therefore framed to investigate these aims: 1 Does accuracy in the use of two functions of the English article system improve over a 10-month period as a result of WCF? 2 Does accuracy in the use of these two functions of the English article system vary according to the WCF options provided?

Design

Context The study was conducted in the English Language Department of a university in Auckland, New Zealand. Students who were new to the university were assigned to a proficiency level after taking a standardized grammar test, a writing test, and a one-on-one interview while students who had previously been studying at a lower proficiency level were placed in the low-intermediate level on the basis of earlier competency-based assessments. The English Language Department describes its approach to the teaching of English as communicative and gives an equal focus to reading, writing, speaking, and listening. Grammatical aspects of the language are also explicitly taught in the programme. The two functions targeted in the study are taught at elementary level but the majority of participants entered the department’s classes at low-intermediate level so there was no guarantee that all participants would have received prior instruction in the use of the article functions.

Participants Most of the students were migrants who had settled in New Zealand within 18 months of commencing study at the low-intermediate level. The students (19 males and 33 females) were predominantly from East Asian countries: Korea (15 per cent), Japan (11 per cent), P.R. China (18 per cent). Other

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Accuracy in using two functions of the English article system was measured over a 10-month period by means of a pre-test–post-test design (a pre-test at the beginning of the 10-month period and post-tests after 2 weeks, 2, 6 and 10 months). Fifty-two low-intermediate ESL learners were arbitrarily assigned to one of four groups: group 1 received direct error correction above each targeted error as well as written and oral meta-linguistic explanation; group 2 received direct error correction and written meta-linguistic explanation; group 3 received direct error correction; group 4 was the control group and therefore did not receive WCF. Each group comprised of 13 students. The study examined the effect of these particular feedback options (variable combinations) as they are well-established practices used by ESL classroom teachers. In adopting this aspect of the design, we acknowledge the need for researchers also to examine the relative effects of single feedback variables.

J. BITCHENER and U. KNOCH

201

countries represented were Vietnam, Yemen, Russia, Switzerland, Saudi Arabia, Chile, Brazil, Serbia, Turkey, Somalia, Romania, Iran, Sri Lanka, India, and Indonesia. The average age of the students was 31.7 years. The majority (78 per cent) claimed to have had formal instruction though their length of earlier study varied across a 7-year period. Essentially equal numbers of participants in the four groups came from backgrounds where articles are included in their L1.

Target structures

Treatment Group 1 received direct error correction above each functional error as well as written and oral meta-linguistic explanation. Direct error correction involved placing a tick or check above correct uses of the two functions, correcting incorrect uses with ‘a’ or ‘the’ above each error, and inserting ‘a’ or ‘the’ where they were omitted but required because of the learner’s noun choice. In the following example, it can be seen that the learner’s first reference to the wind-break was not preceded by the required inclusion of ‘a’ for first mention and in the second reference to the same noun by the required inclusion of ‘the’ for a subsequent mention of the noun: Two people are sitting inside windbreak and are looking over the top of windbreak.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Compared with earlier studies on the value of WCF (Ferris 2002, 2003, 2006), where sometimes as many as 15 linguistic forms and structures had been examined, this study investigated the effect of targeting two functional uses of the English article system: the referential indefinite article ‘a’ for referring to something the first time (first mention) and the referential definite article ‘the’ for referring to something already mentioned (subsequent mention). The decision to limit the focus was based on the positive results that have been found in SLA studies where intensive oral corrective feedback has successfully targeted a single linguistic category (e.g. Carroll and Swain 1993; Muranoi 2000; Ellis et al. 2006 amongst others). The chosen functions were targeted because students across English language proficiency levels (including particularly those at a low-intermediate level of proficiency) experience difficulty in the use of the English article system (Master 1995; Butler 2002; Ferris 2002, 2006; Bitchener et al. 2005). For example, they may experience difficulty deciding whether an article is required and, if it is required, whether it should be the definite or indefinite article. Accuracy in the use of these functions in the pre-test revealed a mean score of 59.41 per cent, thereby indicating that students at a low-intermediate level have only a partial mastery of the functions.

202 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

Because there are occasions when the definite article is required for referring to something for the first time (e.g. uniquely identifiable referents like ‘Pass the butter, please’) or for referring to mass nouns (e.g. ‘People were buying ice-cream at the kiosk’), WCF was not provided on such occasions. A post hoc analysis of 25 per cent of the analysed texts revealed that only 8 per cent of occasions in which articles were used were non-targeted uses. The written meta-linguistic explanation included a simple explanation of the two targeted functional uses of the definite and indefinite articles together with an example of their use. Attached to their pre-test pieces of writing, the students received the following explanation and illustration.

Example A man and a woman were sitting opposite to me. The man was British but I think the woman was Australian. Attention was drawn to this information at each point in the student’s writing where such errors were made. An asterisk referred the students to the meta-linguistic explanations. Oral meta-linguistic explanation took the form of a 30-minute mini-lesson. During this lesson, the researcher explained the meta-linguistic information (rules and example) attached to the students’ returned text. Additional examples were illustrated on the whiteboard and discussed with the class. The students were then given a short ‘controlled practice’ exercise (Appendix 1) and asked to fill the gap in each sentence with ‘a’, ‘the’, or neither. They were given 5 minutes to complete the exercise. The lesson concluded with a plenary discussion of the answers. Group 2 received direct error correction above each functional error and written meta-linguistic explanation. Group 3 only received direct error correction above their errors. Group 4 did not receive feedback. Feedback was only provided in the treatment session that took place a week after the pretest, that is, on the same day as the immediate post-test.

Instruments Each of the five pieces of writing required a description of what was happening in a picture of a social gathering (a beach; a picnic; a campsite; a family celebration, and a sporting event). Picture descriptions were chosen because the range of people, objects, and activities illustrated had the potential to create opportunities for the use of both English article functions. Although, it was acknowledged that they would be able to avoid such uses, if they were uncertain about which was appropriate, and choose other determiners such as ‘one’, ‘two’, ‘this’, ‘that’, it was believed that this would not happen or be possible

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

1 Use ‘a’ when referring to something for the first time. 2 Use ‘the’ when referring to something that has already been mentioned.

J. BITCHENER and U. KNOCH

203

Procedure One week prior to the pre-test, and in accordance with the requirements of the university’s Ethics Committee, the students and the teachers (in separate sessions) in the four classes were provided with information sheets about the study and were given the opportunity to ask questions before signing a participant consent form. Because of the amount of marking required for each class, the data collection took place at different times during the year. On day one, the pre-test was administered. One week later, the treatment (WCF) was provided. For group 1, the immediate post-test was completed after the students had been given 5 minutes to consider the error corrections and the written meta-linguistic explanation and had received the 30-minute lesson (oral meta-linguistic explanation). For group 2, the immediate post-test was completed after the students had been given 5 minutes to consider the error corrections and the written meta-linguistic explanation. For group 3, the immediate post-test was completed after the students had been given 5 minutes to consider the error corrections. For group 4, the immediate post-test was completed as soon as the uncorrected pre-test piece of writing had been returned. The immediate post-test for all groups was returned 1 week after it had been written. Corrective feedback was not provided on this occasion. The first delayed post-test was administered in week eight. The students were not told when the researcher would be returning to the class. The reason for this was to eliminate the possibility of any student studying the feedback that they had earlier been given prior to the test. The researcher did not want the students to be primed in any way beforehand. The first delayed post-test

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

in all linguistic environments. In each of the five writing tasks, no student made fewer than six uses of ‘a’ and ‘the’. Pictures were also chosen because, in ESL classrooms, they are seen as tasks that approximate authentic communication activities. Because the students were at a low-intermediate level of proficiency, some of the key vocabulary items (concrete nouns) were provided around the margins of each picture with arrows pointing to the relevant person, object, or activity. It was decided that this would lower the anxiety level for the students if unknown key words were provided. Additionally, they were allowed to use dictionaries and ask the researcher for a particular word if necessary. Thirty minutes was given for the writing of each description. One of the criticisms that has been levelled against earlier studies that have used a range of instruments and writing conditions concerns the reliability and validity of the data (Ferris 2004; Guenette 2007). Heeding the advice that has been offered so that design flaws such as these are not repeated, this study collected data from genres and writing conditions that were identical across the five writing occasions.

204 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

was returned to the students 1 week later. The second delayed post-test occurred after 6 months and the final delayed post-test occurred after 10 months.

Analysis

RESULTS Table 1 shows the descriptive statistics for the mean test scores for the three treatment groups and the control group at the five different testing periods. Figure 1 provides a visual representation of the mean percentages for the five testing periods for each group. As can be seen, whilst the four groups were very similar at Time 1 (the pre-test), the three treatment groups increased

Table 1: Descriptive statistics for mean test scores by group and testing period Group

N Time 1 Mean SD

1. CF, written, oral 2. CF, written 3. CF 4. Control

Time 2

Time 3

Time 4

Time 5

Mean SD

Mean SD

Mean SD

Mean SD

13 62.46 11.98 89.23

8.22 79.61 10.29 84.31 12.33 88.38

9.53

13 55.31 20.08 83.77 10.36 78.46 11.04 81.69 10.93 88.77 8.55 13 59.69 20.45 79.15 14.27 79.84 13.74 78.31 13.41 81.46 13.90 13 60.17 17.58 67.08 21.45 56.62 22.29 62.46 18.97 58.92 16.16

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Uses of the targeted features were first identified and corrected for each text on each of the five testing occasions. For the texts of students in group 4, the control group, this was done on a photocopy of their writing so that they did not receive the targeted feedback. The same process occurred for all groups with each of the delayed post-tests. Accuracy on each occasion was calculated as a percentage of correct usage. For example, in any one script, three correct uses of the targeted features from 10 obligatory occasions meant a 30 per cent accuracy rate. Inter-rater reliability calculations with a trained research colleague revealed a 95 per cent agreement on the identification of targeted errors and a 98 per cent agreement on the assignment of errors to the targeted categories. Descriptive statistics for the pre-test and the four post-tests were calculated separately for the four groups. Because, no statistically significant differences on the pre-test scores were found, a two-way repeated measure ANOVA was chosen to address the research questions. One-way ANOVAs with Tukey’s post hoc pair-wise comparisons were used to isolate the exact points in time where differences between the groups occurred.

J. BITCHENER and U. KNOCH

205

Effectiveness of feedback types over time 90 85 80 CF, written and oral CF, written CF

75 70 65

Control

60 55 Time 1

Time 2

Time 3

Time 4

Time 5

Figure 1: Mean percentage for five testing periods for each group

Table 2: Two-way ANOVA analysis Source Between subjects WCF type Within subjects Time Time X WCF type

df

F

p

3

6.21

0.001

4 12

37.21 5.95

<0.001 <0.001

their mean scores at Time 2 (the immediate post-test). The control group also increased its mean score at Time 2, but to a much lesser extent than the three treatment groups. In order to compare the treatment and control groups’ test scores, a series of ANOVAs were computed. Because a one-way ANOVA indicated no statistically significant difference between the four groups, F(3, 48) = 0.525, p = 0.667, a two-way repeated measures ANOVA was chosen to address the research questions. In this two-way ANOVA, the test scores were entered as the dependent variable with Time (five levels) and WCF type (four levels) as independent variables. Table 2 shows the results of the analysis. As can be seen in Table 2, there was a significant interaction between Time and WCF type, indicating that the different groups developed differentially over time. To statistically examine the differences between pairs of groups, post hoc multiple comparison tests were performed. One-way ANOVAs revealed that the differences in the scores were significant at Time 2 [F(3, 48) = 5.515; p = 0.002], Time 3 [F(3, 48) = 7.350; p = 0.000],

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

50

206 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

Time 4 [F(3, 48) = 6.156; p = 0.001], and Time 5 [F(3, 480) = 16.602; p = 0.001]. Tukey’s post hoc pair-wise comparison (with an alpha level of .05) was performed to isolate the significant differences among the three groups. These indicated that at all four post-tests (Time 2–Time 5), all treatment groups outperformed the control group and that there were no statistically significant differences between the treatment groups.

DISCUSSION Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Considering the first research question, all three groups that received WCF outperformed the group that received no feedback in all four post-tests even though all groups developed differently over time. This result corroborates those of several recent studies on article use (Sheen 2007; Bitchener 2008; Bitchener and Knoch 2008a, 2008b, 2009) that examined the effect of WCF on new pieces of writing over a 2-month period. The enduring effect on accuracy over a 10-month period is clear evidence of the role that WCF can play in helping learners raise the level of accuracy with which they use simple linguistic forms/structures like the two functional uses of the English article system targeted in this study. It demonstrates the value of this aspect of explicit learning for developing at least explicit declarative knowledge that can be retained over time. Until recently, the focus in WCF research has been on text revision. The special contribution of the findings of this study to the on-going debate about whether WCF is effective or not is that it demonstrates the potential for WCF to have a positive effect on the acquisition of simple linguistic forms/structures like those targeted in the study—a role that goes beyond that of helping learners accurately revise their texts. If only pseudo-learning had occurred, the effectiveness of the treatment would not have been as durable. As such, it brings into question Truscott’s (1996) claim that WCF is ineffective but, on the other hand, supports his qualification that there may be some special cases where WCF might have value. Whereas earliest research has focused on a comprehensive range of error categories, the results of this study demonstrate the value of focusing on a single-error category. In this respect, they corroborate the findings of several recent studies on WCF (Bitchener et al. 2005; Sheen 2007; Bitchener 2008; Bitchener and Knoch 2008a, 2008b, 2009) and on oral corrective feedback (Doughty and Varela 1998; Muranoi 2000; Han 2002). Because this study focused on two functional uses of the English article system, further research is now required to determine the extent to which WCF is effective in helping learners acquire other forms/structures that they use incorrectly. It is especially important that it be tested with more complex features to determine whether or not its optimal effect is with single rule-based functions such as those examined in this study.

J. BITCHENER and U. KNOCH

207

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

With respect to the second research question, the observable differences in effect for the three different WCF options in the four post-tests were found to not be statistically significant. At Time 5, it appears that there may have been a slight advantage for the meta-linguistic groups over the error correction group only, but this observed difference was not statistically significant. Another interesting aspect of the performance of the meta-linguistic groups occurs at Time 3. In this delayed post-test, their accuracy levels fell. A post hoc review of the students’ scripts revealed a higher article overuse— a phenomenon that can sometimes arise when meta-linguistic feedback or instruction has just been provided. Like earlier findings (Bitchener 2008; Bitchener and Knoch 2008a, 2008b, 2009), the study revealed that the simple provision of error correction was just as effective as the additional provision of written and oral meta-linguistic explanation. Although Bitchener (2008) found that only two of the three treatment groups outperformed the control group (there was no significant difference between participants who received direct error correction with written meta-linguistic explanation and those who received no feedback), the observed differences between the three treatment groups was found to not be statistically significant. This may have been an artefact of sample size (n = 75, in Bitchener 2008; and n = 144, in Bitchener and Knoch 2008a). Differences in effect were evident over time (3 and 2 months, respectively) in two other studies (Bitchener et al. 2005; Sheen 2007) between groups that received meta-linguistic explanation and those that received direct error correction. In Sheen’s study, the difference was evident in the delayed post-test but not in the immediate post-test. The extent to which the students in each study sought additional input outside the classroom between the post-test periods may be a reason for the difference. Differences in type, amount and delivery of meta-linguistic explanation between the two studies might also explain the improvement reported in Sheen’s study. That no advantage was evident for any of the treatment groups in the present study on any of the four post-test occasions suggests that as long as some form of WCF is provided, it may have a positive effect on accuracy and the acquisition of simple functional uses of the English article system at least. It is interesting to note that the additional time on task for the group that received the oral meta-linguistic feedback during the mini lesson did not reveal an advantage for this group. Further research is needed to see if there is an advantage for different types, amount, frequency, and delivery of metalinguistic explanation over a range of testing occasions. Additionally, further research is needed to determine whether or not written meta-linguistic explanation is more beneficial than oral meta-linguistic explanation and whether or not meta-linguistic explanation has an advantage over other types of direct feedback across a range of linguistic error categories. It may be the case, as was demonstrated in Bitchener et al. (2005) that system or rule-based forms/ structures like the use of article functions are more amenable to WCF or particular types of WCF than item-based features like prepositions, for example.

208 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

CONCLUSION

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The aims of this study were to investigate the efficacy of WCF on targeted linguistic error categories (two functional uses of the English article system) in new pieces of writing over time and to determine whether or not there was a differential effect on accuracy for three direct corrective feedback options (direct error correction with written and oral meta-linguistic explanation; direct error correction with written meta-linguistic explanation; direct error correction). In targeting two functional uses of the English article system (the indefinite article ‘a’ for first mention and the definite article ‘the’ for subsequent mentions), the study found that learners who received WCF outperformed those who received no feedback and that this pattern continued over a 10-month period. Concerning the second aim, the study found no difference in effect between the three treatment groups, indicating, therefore, that any one option was as effective as another. The contribution that this study makes to the existing literature is its demonstration of the role that WCF can have on the long-term acquisition of certain linguistic forms/structures and, as such, supports Truscott’s prediction that WCF may be effective for helping learners acquire some linguistic forms/structures. Earlier studies have identified its value for helping learners improve the accuracy of their writing when undertaking text revisions and several recent studies have demonstrated its role in helping learners overcome certain recurring linguistic errors. No study that we are aware of has sought to investigate its effect over such an extensive period of time. We acknowledge that the study only considered the effect of WCF on the use of two functions of the English article system but given its positive findings we would suggest that further research investigate its effect on other linguistic error categories. Also, because the study focused on partially known linguistic forms/structures, further research is needed to examine the extent to which it might be able to facilitate the acquisition of those that learners have not been previously taught. A further theoretical contribution of the study to existing knowledge is its discovery that there was no advantage for any one of the direct feedback options. This does not mean, however, that certain variables (separated feedback types as opposed to combinations of feedback types) may or may not have a differential effect on accuracy development. Further research is required to determine whether the advantage reported by Sheen (2007) for meta-linguistic explanation is retained over a more extensive period of investigation and to determine whether type, frequency, amount, and delivery of meta-linguistic explanation are factors in any difference. Singlestudy investigations of any difference between direct and indirect corrective feedback options are also needed to resolve the conflicting findings of earlier comparative studies. This line of investigation is all the more germane

J. BITCHENER and U. KNOCH

209

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

given the positive findings that are now emerging on the long-term effects of WCF. Inevitably, some limitations need to be acknowledged. Because of the difficulty in accessing participants over an extensive period of time, the sample size, while acceptable, was smaller than one would have wished for. Second, the study was interested in examining the effect of typical WCF practices in second language classrooms, practices that on some occasions involve only direct error correction, practices that on other occasions add written meta-linguistic explanation and or oral meta-linguistic explanation. Consequently, it is not possible to determine if, in cases where direct error correction together with written and/or oral meta-linguistic explanation are provided, the findings can be attributed to the effect of one or more of the feedback variables. Consequently, studies that separate the feedback variables are now required to determine any relative effect. Third, the use of the same type of writing task on five occasions might also be considered a limitation because it does not include opportunities for learners to demonstrate their ability to perform with the same level of accuracy when writing in other genres. Future researchers might, therefore, wish to include such opportunities in at least one of their delayed post-tests. It could also be argued that learners are primed when asked to do a writing task like those they have already done. While every effort was made to preserve ecological validity by the manner in which the investigation was conducted in the classroom, it could, nevertheless, be argued that similar results may be difficult to achieve in real classroom situations where individual learners needs and rates of progress are considered important. For this reason, we recommend that further research be undertaken to determine the effectiveness of WCF for treating a range of linguistic forms/structures when learners are grouped according to common errors being made and when they are given individual feedback. In order to track individual and group progress, students could be encouraged to keep logs of their progress. Case study reports on these types of investigation would complement current research approaches and thereby provide richer insights into the effectiveness of WCF. Despite these limitations, a number of pedagogical recommendations can be offered. Teachers should feel confident about providing direct WCF on their students’ errors in the use of the two functions of the English article system. We would suggest that a targeted approach for a period of time be adopted (until there are clear signs of accuracy improvement) and that negotiation with learners take place to ensure that they agree about what error category is going to be focused on. This study has shown that a single feedback session can be effective in developing accuracy in the use of two rule-based features but if teachers are able to provide additional feedback on more occasions, it is possible that the accuracy rate may

210 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

increase and that the amount of time required to achieve a high level of mastery may be reduced. The study has also shown that the provision of error correction alone on specific rule-based functions may be just as effective as combining it with written and oral meta-linguistic explanation so, in similar contexts, teachers may not necessarily feel that they need to go beyond this approach. Finally, we would recommend that further research investigate the extent to which different types of direct and indirect WCF enable learners to acquire over time other forms/structures that prove to be problematic.

Fill the gaps with a (an), the, or neither. 1 This morning I bought ______ newspaper and ______ magazine. ______ newspaper is in my bag but I don’t know where I put ______ magazine. 2 I saw ______ accident this morning. ______ car crashed into ______ tree. ______ driver of ______ car wasn’t hurt but ______ car was badly damaged. 3 There are two cars parked outside: ______ blue one and ______ grey one. ______ blue one belongs to my neighbours; I don’t know who ______ owner of ______ grey one is. 4 My friends live in ______ old house in ______ small village. There is ______ beautiful garden behind ______ house. I would like to have ______ garden like that.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

APPENDIX 1: CONTROLLED PRACTICE EXERCISE

211 J. BITCHENER and U. KNOCH

APPENDIX 2: INSTRUMENT EXAMPLES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

212 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

REFERENCES DeKeyser, R. 2004. ‘Implicit and explicit learning’ in A. Davies and C. Elder (eds): The Handbook of Applied Linguistics. Blackwell Publishers, pp. 313–48. Doughty, C. and E. Verela. 1998. ‘Communicative focus on form’ in C. Doughty and J. Williams (eds): Focus on Form in Classroom Second Language Acquisition. Cambridge University Press, pp. 114–38. Ellis, R. 2004. ‘The definition and measurement of L2 explicit knowledge,’ Language Learning 52/2: 227–75. Ellis, R., S. Loewen, and R. Erlam. 2006. ‘Implicit and explicit corrective feedback and the acquisition of L2 grammar,’ Studies in Second Language Acquisition 28: 339–68. Fathman, A. and E. Whalley. 1990. ‘Teacher response to student writing: Focus on form versus content’ in B. Kroll (ed.): Second Language Writing: Research Insights for the Classroom. Cambridge University Press, pp. 178–90. Ferris, D. R. 1995. ‘Can advanced ESL students be taught to correct their most serious and frequent errors?’ CATESOL Journal 8: 41–62. Ferris, D. R. 1997. ‘The influence of teacher commentary on student revision,’ TESOL Quarterly 31: 315–39. Ferris, D. R. 1999. ‘The case for grammar correction in L2 writing classes. A response to Truscott (1996),’ Journal of Second Language Writing 8: 1–10. Ferris, D. R. 2002. Treatment of Error in Second Language Student Writing. University of Michigan Press. Ferris, D. R. 2003. Response to Student Writing: Implications for Second Language Students. Lawrence Erlbaum Associates. Ferris, D. R. 2004. ‘The ‘‘Grammar Correction’’ debate in L2 writing: Where are we, and where do we go from here? (and what do we do in the meantime . . . .?),’ Journal of Second Language Writing 13: 49–62. Ferris, D. R. 2006. ‘Does error feedback help student writers? New evidence on the shortand long-term effects of written error correction’ in K. Hyland and F. Hyland (eds): Perspectives on Response. Cambridge University Press.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Ashwell, T. 2000. ‘Patterns of teacher response to student writing in a multi-draft composition classroom: Is content feedback followed by form feedback the best method?’ Journal of Second Language Writing 9/3: 227–57. Bitchener, J. 2008. ‘Evidence in support of written corrective feedback,’ Journal of Second Language Writing 17: 102–18. Bitchener, J. and U. Knoch. 2008a. ‘The value of written corrective feedback for migrant and international students,’ Language Teaching Research Journal 12: 409–31. Bitchener, J. and U. Knoch. 2008b. ‘The value of a focussed approach to written corrective feedback,’ ELT Journal available at http://eltj. oxfordjournals.org/cgi/content/full/ccn043v1. Last accessed 13 August 2008. Bitchener, J. and U. Knoch. 2009. ‘The relative effectiveness of different types of direct written corrective feedback,’ System 37: 322–29. Bitchener, J., S. Young, and D. Cameron. 2005. ‘The effect of different types of corrective feedback on ESL student writing,’ Journal of Second Language Writing 9: 227–58. Butler, Y. 2002. ‘Second language learners’ theories on the use of English articles,’ Studies in Second Language Acquisition 24: 451–80. Carroll, S. 2001. Input and Evidence: The Raw Material of Second Language Acquisition. Benjamins. Carroll, S. and M. Swain. 1993. ‘Explicit and implicit negative feedback: an empirical study of the learning of linguistic generalizations,’ Studies in Second Language Acquisition 15: 357–86. Chandler, J. 2000. ‘The efficacy of error correction for improvement in the accuracy of L2 student writing’. Paper presented at the AAAL Conference, Vancouver, B.C., March. Chandler, J. 2003. ‘The efficacy of various kinds of error feedback for improvement in the accuracy and fluency of L2 student writing,’ Journal of Second Language Writing 12: 267–96. Clahsen, H., J. Meisel, and M. Pienemann. 1983. Deutsch als Zweitsprache. Der Spracherwerb auslandischer Arbeiter. Narr. DeKeyser, R. 1993. ‘The effect of error correction on L2 grammar knowledge and oral proficiency,’ Modern Language Journal 77: 501–14.

J. BITCHENER and U. KNOCH

Lalande, J. F. 1982. ‘Reducing composition errors: An experiment,’ Modern Language Journal 66: 140–49. Leeman, J. 2003. ‘Recasts and second language development: Beyond negative evidence,’ Studies in Second Language Acquisition 25: 37–63. Lyster, R. 2004. ‘Differential effects of prompts and recasts in form-focused instruction,’ Studies in Second Language Acquisition 26: 399–432. Mackey, A. and R. Oliver. 2002. ‘Interactional feedback and children’s L2 development,’ System 30: 459–77. Mackey, A. and J. Philp. 1998. ‘Conversational interaction in second language development: Recasts, responses, and red herrings?’ The Modern Language Journal 82: 338–56. Mackey, A., J. Philp, T. Egi, A. Fujii, and T. Tatsumi. 2002. ‘Individual differences in working memory, noticing of interactional feedback and L2 development’ in P. Robinson (ed.): Individual Differences and Instructed Language Learning. John Benjamins, pp. 181–210. Master, P. 1995. ‘Consciousness raising and article pedagogy’ in D. Belcher and G. Braine (eds): Academic Writing in a Second Language: Essays on Research and Pedagogy. Ablex Publishing Corporation. McDonough, K. 2006. ‘Interaction and syntactic priming: English L2 speakers’ production of dative constructions,’ Studies in Second Language Acquisition 28: 179–207. Muranoi, H. 2000. ‘Focus on form through interaction enhancement: Integrating formal instruction into a communicative task in EFL classrooms,’ Language Learning 50: 617–73. Nagata, H. 1993. ‘Intelligent computer feedback for second language instruction,’ Modern Language Journal 77: 330–39. Pienemann, M. 1998. Language Processing and Second Language Development: Processability Theory. Benjamins. Polio, C., N. Fleck, and N. Leder. 1998. ‘ ‘‘If only I had more time’’: ESL learners’ changes in linguistic accuracy on essay revisions,’ Journal of Second Language Writing 7: 43–68. Robb, T., S. Ross, and I. Shortreed. 1986. ‘Salience of feedback on error and its effect on EFL writing quality,’ TESOL Quarterly 20: 83–93.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Ferris, D. R. and M. Helt. 2000. ‘Was Truscott right? New evidence on the effects of error correction in L2 writing classes’. Paper presented at the American Association of Applied Linguistics Conference, Vancouver, B.C., March. Ferris, D. R. and B. Roberts. 2001. ‘Error feedback in L2 writing classes: How explicit does it need to be?,’ Journal of Second Language Writing 10: 161–84. Ferris, D. R., S. J. Chaney, K. Komura, B. J. Roberts, and S. McKee. 2000. ‘Perspectives, problems, and practices in treating written error’. Colloquium presented at International TESOL Convention, Vancouver, B.C., March. Guenette, D. 2007. ‘Is feedback pedagogically correct? Research design issues in studies of feedback on writing,’ Journal of Second Language Writing 16: 40–53. Han, Z. 2002. ‘A study of the impact of recasts on tense consistency in L2 output,’ TESOL Quarterly 36: 543–72. Havranek, G. and H. Cesnik. 2003. ‘Factors affecting the success of corrective feedback’ in S. Foster-Cohen and A. Nizegorodzew (eds): EuroSLA Yearbook, vol. 1. Benjamins. Hendrickson, J. 1981. Error Analysis and Error Correction in Language Teaching. SEAMEO Regional Language Centre. Iwashita, N. 2003. ‘Negative feedback and positive evidence in task-based interaction: Differential effects on L2 development,’ Studies in Second Language Acquisition 25: 1–36. Kepner, C. G. 1991. ‘An experiment in the relationship of types of written feedback to the development of second-language writing skills,’ Modern Language Journal 75: 305–13. Kim, H. and G. Mathes. 2001. ‘Explicit vs. implicit corrective feedback,’ The Korea TESOL Journal 4: 1–15. Krashen, S. 1982. Principles and Practice in Second Language Acquisition. Prentice-Hall. Krashen, S. 1985. The Input Hypothesis. Longman. Krashen, S. 1994. ‘The input hypothesis and its rivals’ in N. Ellis (ed.): Implicit and Explicit Learning of Languages. Academic Press, pp. 45–77. Krashen, S. 1999. ‘Seeking a role for grammar: A review of some recent studies,’ Foreign Language Annals 32: 245–57.

213

214 THE CONTRIBUTION OF WCF TO LANGUAGE DEVELOPMENT

Swain, M. 2005. ‘The output hypothesis; theory and research’ in E. Hinkel (ed.): Handbook on Research in Second Language Teaching and Learning. Lawrence Erlbaum, pp. 471–84. Truscott, J. 1996. ‘The case against grammar correction in L2 writing classes,’ Language Learning 46: 327–69. Truscott, J. 1999. ‘The case for ‘‘the case for grammar correction in L2 writing classes’’: A response to Ferris,’ Journal of Second Language Writing 8: 111–22. Truscott, J. 2004. ‘Dialogue: Evidence and conjecture on the effects of correction: A response to Chandler,’ Journal of Second Language Writing 13: 337–43. Truscott, J. 2007. ‘The effect of error correction on learners’ ability to write accurately,’ Journal of Second Language Writing 16: 1–18.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Rosa, E. and R. Leow. 2004. ‘Awareness, different learning conditions, and second language development,’ Applied Psycholinguistics 25: 269–92. Schmidt, R and S. Frota. 1986. ‘Developing basic conversational ability in a second language: A case study of an adult learner of Portuguese’ in R. Day (ed.): Talking to Learn: Conversation in Second Language Acquisition. Newbury House, pp. 237–326. Semke, H. 1984. ‘The effects of the red pen,’ Foreign Language Annals 17: 195–202. Sheen, Y. 2007. ‘The effect of focused written corrective feedback and language aptitude on ESL learners’ acquisition of articles,’ TESOL Quarterly 41: 255–83. Sheppard, K. 1992. ‘Two feedback types: Do they make a difference?’ RELC Journal 23: 103–10.

Applied Linguistics: 31/2: 215–235 ß Oxford University Press 2009 doi:10.1093/applin/amp023 Advance Access published on 18 June 2009

Probabilities and Surprises: A Realist Approach to Identifying Linguistic and Social Patterns, with Reference to an Oral History Corpus ALISON SEALEY

The relationship between language and identity has been explored in a number of ways in applied linguistics, and this article focuses on a particular aspect of it: self-representation in the oral history interview. People from a wide range of backgrounds, currently resident in one large city in England, were asked to reflect on their lives as part of a project to celebrate the millennium, resulting in a corpus of 144 transcribed interviews. The article considers the utility of realist social theory and complexity theory in the analysis of patterns—and deviations from those patterns—in both the linguistic features of these interviews and the social categories to which people are routinely ascribed. Corpus linguistic software was used to identify discourse features of the corpus as a whole, and to compare and contrast features produced by different speakers with reference to the conventional social categories used in quantitative research. These categories, with their homogenizing limitations, are challenged with reference to complex causation. The article uses the category of gender to exemplify the multi-method approach advocated.

INTRODUCTION This article is concerned with probabilistic patterns, including deviations from them, in social behaviour in general and in language use in particular. The approach conceives of social and linguistic processes as complex, agencydriven, and susceptible to changing kinds of analysis as computer technology develops. The data used to illustrate the discussion are a corpus of 1.8 million words of transcribed speech, comprising 144 oral history interviews. As examples of discourse, they constitute a rich resource for exploring the probabilistic—but not determined—linguistic patterns to be found in a set of texts that have much in common with each other but that are each nevertheless unique; each interviewee demonstrates the ever-present potential for linguistic creativity while simultaneously contributing to the collective entity that emerges as ‘the discourse of life histories’. Located in between the composite identity of the whole set and the unique identity of the individual interview are some patterns associated with the speakers’ membership of various sub-categories, identifiable from both social and linguistic analysis.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

University of Birmingham, UK

216 PROBABILITIES AND SURPRISES

The data The interviews which comprise the data in this study were recorded in 2000– 2001 by two oral historians as part of the ‘Millennibrum’ Project and deposited in the Local Studies and History section of Birmingham Central Library.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The original aim of the project within which these interviews were generated was not specifically to contribute to linguistic research. Rather, it was to record, from a socio-cultural perspective, the accounts of a diverse range of residents of the large English city of Birmingham, in order to recognize and celebrate the experiences of its citizens. From this angle, the corpus represents a potential means of understanding complex social processes, where, again, each individual is a social actor contextualized in a complex web of structured social relations. As Uprichard and Byrne (2006: 668) put it, ‘narratives are descriptions not of single systems but of the interweaving of complex systems,’ because, ‘[p]eople never tell just the story of their own life; nor do they project simply in terms of themselves. All lives are embedded in the social; there is no personal without the social’. The approach taken here aims to bring together insights from both realist social theory and applied corpus linguistics. From the former comes an emphasis on the duality of structure and agency, along with the recognition that while social actors have interests and intentions, their scope for realising these is constrained. As Archer (2000: 262) expresses it, ‘Because of the preexistence of those structures which shape the situation in which we find ourselves, they impinge upon us without our compliance, consent or complicity’. From the latter comes the recognition that in making meaning in speech and writing, human beings are both constrained and enabled by the linguistic choices available to them. That is, in the formulation closely associated with Halliday (e.g. 1989, 1991), instances of discourse represent choices from within the language system, and, as Sinclair (1991) suggests, speakers deploy ready-made sequences (the ‘idiom principle’) but remain able to exercise ‘open choice’ as well—although if speakers are to be readily understood, they are constrained to choose from within pre-existing grammatical and semantic resources. These observations are consistent with a view of the social, discursive world as systematic—patterned and often predictable, but where the systems in play are open and dynamic, with human meanings and human agency not only reproducing familiar patterns, but also generating novelty and surprise. The approach to research explored in this article is concerned with both trends and probabilities on the one hand, and variation on the other, in relation to both biographical experience and the discursive representation of that experience in the corpus of 144 transcribed oral history interviews. Which people, with which kinds of attributes, use language in similar/ contrasting ways to tell their life histories? The article concludes with a case study of a sub-group identified by the method presented.

A. SEALEY

217

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The aim was to preserve the narrative accounts of a diverse range of residents of the city at the turn of the millennium, for local people to participate ‘in presenting and recording their experiences, beliefs, contributions to the community and hopes for the future’ (Dick 2002). Of these 150 interviews, each lasting up to 90 minutes, 144 have been released for further research in accordance with the ethical consent procedures employed by the library. For each interviewee, information is available about place of birth, age, sex, occupation, level of education, marital status, and religious affiliation—all topics which are usually covered in the interviews themselves. The interviews were transcribed as part of the original project, in which I was not directly involved, but (with the support of a small grant) I oversaw the post-editing of the texts to make them suitable for corpus analysis. This included checking samples of the transcripts against the sound recordings, standardizing transcript conventions, indicating speaker turns, so that the (usually brief) contributions of the interviewer could be excluded from the analysis when desirable, and anonymizing the interviews and the spreadsheet of meta-data which records demographic details such as age, sex, place of birth, and so on. The oral historians who conducted the interviews are skilled in encouraging people to articulate their memories, views, and beliefs. The interviews were conducted, by one of the two interviewers, usually in the interviewees’ homes, lasting up to 90 minutes, so that the full corpus represents about 250 hours of recording. During the conversation, both interviewer and interviewee held a postcard, with a few single-word subject headings on it. The topics covered in the interviews varied slightly, but generally included, to reflect the funders’ goals for this project: the interviewees’ childhood memories and experiences of school; first experiences of work and subsequent jobs; family life before leaving home, and relationships with parents and siblings; adult relationships, including courtship, marriage and, in some cases, the breakdown of relationships; experiences of moving and migration; parenting and hopes for their children’s future. In addition, interviewees were asked about the role of religion in their lives, about social changes they had witnessed in their own lifetimes, and particularly about changes in Birmingham they had noticed during the time they had lived in the city. All had the option, of course, to omit any of these topics from their account if they wished. Despite the similarities in themes, each interview is a record of a specific social interaction, and each interviewee interprets this in his or her own way. Interviewees inevitably make judgements about the interviewer and her expectations, including about how far she shares their knowledge about the things they reference. For example, interviewees who are much older than their interlocutor, or who have lived in places beyond Birmingham, tend to assume that some of their experiences will be unfamiliar to the interviewer, and so they explain them in greater detail. In addition, these interviews cannot be neutral descriptions, or representations, of each ‘self’ and its history, as they are interactional tellings, produced in a context of

218 PROBABILITIES AND SURPRISES

interpretation and negotiation (Wortham 2000; Pavlenko 2007). Nevertheless, there is sufficient homogeneity about the interviews for them to have certain features in common, including linguistic features.

THE DISCOURSE OF THE ‘MILLENNIBRUM’ INTERVIEWS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The shared linguistic features of the corpus have been identified using a range of techniques, including software developed specifically for the analysis of electronic corpora, mainly the suite of applications, WordSmith Tools (Scott 2004), and WMatrix (Rayson 2008a). From a purely lexical point of view, the detailed consistency wordlist generated by WordSmith Tools can be consulted to identify those items that are found in all—or the majority—of the 144 transcripts. There are 46 items that occur in every text, of which a large proportion, predictably, comprises grammatical words. Of the rest, all are core vocabulary, but a glimmer of the ‘genre’ of ‘life history’ is suggested by the list: good, home, know, like, old, school, see, still, things, think, time, way. More revealing, perhaps, is the list of frequent n-grams (i.e. sequences of ‘n’ consecutive characters, including spaces, such that recurring ‘chunks’ of text longer than single words can be identified). Strings of three words occurring in 130 or more of the texts include: a lot of/to go to/I went to/and it was/it was a/and I was/when I was/one of the/I don’t know/there was a/and that was/I used to/and I think/I think it/I was born/I had a/but it was/and there was. Strings of four words occurring in 100 or more of the texts include: I think it was/I was born in/and I used to/and we used to. Two features are striking about these lists, both, unsurprisingly, concerned with the articulation of memories. There is a preponderance of expressions concerned with past time, and, in the use of ‘private’ verbs, some hints of tentativeness about relating what happened (‘I think’, ‘I don’t know’). A third way of determining what is distinctive about these texts as a whole is by comparing this corpus with a reference corpus. Different analysts use slightly different calculations to identify items which are statistically significantly more frequent in the target corpus than the reference corpus, but both Scott’s (2004) WordSmith Tools and Rayson’s (2008a) WMatrix use log likelihood (LL) values to generate lists of ‘key’ items. In this operation, ‘the word that has the most significant relative frequency difference between the two corpora’ has the largest LL value, and thus ‘the words most indicative (or distinctive) of one corpus, as compared to the other corpus, occur at the top of the list’ (Rayson 2008b: 528). In the present study, WMatrix was used to identify which items are ‘key’ in comparison with the Spoken Data in the BNC Sampler Corpus. The top 38 items (those with a log likelihood value of more than 450) include: was, and, my, $place (the replacement term, to preserve anonymity, for specific places mentioned in the interviews) used_to, had, school, because, Birmingham, I, were, very, people, me, went, am, mom, remember, in, in_those_days, children, years, family, to, had_to, father, life, always, mother, friends, university, parents,

A. SEALEY

219

road, at_the_time, worked, at, quite, city. The semantic fields connoted by the more content-carrying of these items is fairly self-evident, and accords with the topics chosen for this project, but another way of representing them is using the semantic ‘domain clouds’ available in WMatrix (Figure 1). The WMatrix tool can be used to illustrate the relative frequency differences between the Millennibrum corpus and a reference corpus in a similar manner to the ‘tag clouds’ employed in some social networking web sites. In those, ‘an alphabetically sorted list of words (confusingly for this context called tags) are shown in a larger font if they are (manually) assigned more frequently to shared digital photographs . . . or web site bookmarks . . .’ (Rayson 2008b: 533). WMatrix incorporates the USAS tagger (Rayson et al. 2004, cited in Rayson 2008b), which automatically assigns semantic fields (domains) to each word or multiword expression in a corpus. The clouds it produces use larger fonts to indicate greater keyness, so that, in this case, the semantic domains of ‘moving,_coming_and_going’ and ‘personal_relationship:_ general’, among others, are shown to be significantly more frequent in this corpus than in the spoken component of the BNC. From each of these sets of findings, we can conclude that there is a commonality about these texts, despite the fact that they represent 144 different and, in some ways, strikingly contrasting life stories. Among these interviewees are individuals who are young and old (their years of birth span 1896–1985), migrants from across the world and ‘Brummies born-and-bred’, employees from all the major occupation categories of the SOC2000 classifications (ONC 2000), conventional heterosexual family members and people

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Figure 1: Screenshot of part of the output from the ‘domain cloud’ identification by WMatrix of semantic themes in the oral history corpus compared with the BNC Sampler Spoken Corpus

220 PROBABILITIES AND SURPRISES

with ‘alternative’ lifestyles—and yet all have produced a text in this context which contributes to a patterned discourse with identifiable features.

DISCOURSE AS A COMPLEX DYNAMIC SYSTEM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

One way to account for this relative homogeneity is with reference to complex systems theory, whose application to empirically observed language behaviour has been most thoroughly developed by Larsen-Freeman and Cameron (2008). From this perspective, the production in real time of interactions such as these oral history interviews is accomplished by the speakers within a nested complex of contexts, by means of a potentially extensive set of linguistic resources that are nevertheless constrained in various ways. As outlined elsewhere (Sealey and Carter 2004; Carter and Sealey in press), these interactions can be conceptualized and described from different analytical perspectives. Each speaker has their own ‘psychobiography’ (Layder 1997), the unique complex of experiences—and reflections on those experiences—that are ‘integral to the experiencing individual’, each person’s ‘individual truth’ (Craib 1998: 31). (In the case of these interviews, however, the interviewer takes steps to downplay, or mute her individuality and provide maximum space for the production of the interviewee’s contributions.) A person’s lifespan may be considered in terms of the ‘ontogenetic timescale’ (Larsen-Freeman and Cameron 2008: 168–9), whereas each moment of ‘online talk’ is experienced on the ‘microgenetic timescale’. In the collaborative event that is the interview, which in Layder’s (1997) classification occupies the ‘domain’ of ‘situated interaction’, the interlocutors ‘soft assemble their contribution’, as Larsen-Freeman and Cameron (2008: 169) express it. ‘Soft assembly,’ they explain, ‘describes an adaptive action in which all aspects of context can influence what happens at all levels of activity’. Operating on different timescales still, but influencing each interview—and to some extent acting as components of them—are what Layder (1997) terms the ‘domains’ of social settings and contextual resources, highlighting the socio-historical context of structured social and economic relations. This approach is compatible with Craib’s (1998: 28) acknowledgement of ‘the wider social structure and . . . the wider historical processes which provide us with the stage on which we act out our lives’. Larsen-Freeman and Cameron (2008: 169), similarly, situate interactions within various ‘levels’ of social systems, from the dyad of this kind of interaction, to ‘socio-cultural groups and institutions of various types and sizes all the way up to the society of the speech community’. These insights from realist social theory and complexity theory present us with the beginnings of an explanation for the findings from this corpus analysis of linguistic features that are common to these examples of the life histories collected in one large English city at the turn of the millennium. Out of all the possible human sounds these speakers might have recorded, only a circumscribed sub-set of words, phrases and grammatical constructions, intelligible

A. SEALEY

221

. . . invokes the image of a multidimensional landscape with hills and valleys, over which the system roams, creating a trajectory . . .. The landscape represents the probabilities of various modes or phases of discourse behaviour, and the trajectory is carved out as a particular conversation moves from one mode to another. Crucial to the description I am proposing is the centrality of human agency, where each interviewee, in the context of the complex, dynamic, interactional processes outlined above, makes choices about how to formulate this specific account of their life experience. As Craib (1998: 28) puts it, ‘We are each of us given a starting point and we do something with it.’ If the generic oral life history has certain features, then, in accordance with the fractal patterns typically found in complex systems (Larsen-Freeman and Cameron 2008: 109–111), each individual instance has its own pattern that potentially mimics the overall composite on a smaller scale.

APPROACHES TO ANALYSIS By means of corpus analysis, it is possible to identify both similarities and differences between the whole set of interviews and each individual example, and also to find patterns among sub-sets of the corpus as a whole. In the tradition of quantitative, variationist analysis, the conventional approach involves predetermining which demographic ‘variables’ are likely to correlate with the differential use of particular linguistic features, and to use some kind

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

to speakers of English, was actually drawn on. From within that stock of resources, a further sub-set accounts for all of what was produced, and this can be partially explained with reference to insights from psycholinguistics (e.g. scripts, schemas, and the ‘economic’ advantages of processing expected, rather than novel, input and output; e.g. Rumelhart 1975; 1984) and pragmatics (e.g. the co-operation principle (Grice 1975) and theories of relevance (Sperber and Wilson 1986)); also helpful are sociologically derived understandings about sedimented patterns of social behaviour, with their parallels in corpus linguistic findings about ‘discourse patterns’ that cluster around the ‘meanings [that] are repeatedly expressed in a discourse community,’ and that work by ‘filtering and crystallizing ideas, and by providing prefabricated means by which ideas can be easily conveyed and grasped’ (Stubbs 1996: 158). One indicative finding here is that, from the total vocabulary stock of the English language, just 24,276 items (types) feature in this corpus of just under 1.8 million words (using WordSmith Tools’ Wordlist function, which also calculates the standardized type/token ratio for this corpus as 32.89). In the account I am seeking to provide, the interviews represent examples of discourse as a complex system, which, as Larsen-Freeman and Cameron (2008: 175) suggest:

222 PROBABILITIES AND SURPRISES

of regression analysis to identify correlations. There are several problems with this, however, given the claims made above about the nature of social and linguistic phenomena as complex systems, and the notion of causality implied by the theory deployed here. Models which assume that speaker characteristics are ‘independent variables’, with linguistic features as ‘dependent variables,’ imply a linear model of causality. Such models do not allow for the interaction between the variables, and they do not model well the dynamic, systems-based realities with which we are concerned in this kind of data. The limitations of linear modelling are succinctly summarized by Byrne (2002a: 112):

Methods of analysis that are suitable for data collected in experiments are unlikely to be appropriate for the analysis of linguistic data such as the interview transcripts described above. For example, derived from an experimental paradigm, sex (or gender) would be thought of as a ‘variable’; the interviews of the male speakers would be separated from those of the females and compared with them on various linguistic dimensions. Indeed, Cameron (2005: 484), in the context of advocating a post-modern perspective on gender, maintains that even a ‘modern feminist approach’ to language and gender research ‘presupposes the existence of two internally homogeneous groups, ‘‘men’’ and ‘‘women’’, and looks for differences between them’, and the Keyword tool in WordSmith Tools allows a corpus analyst, should they choose, to identify words which occur with significantly greater frequencies in either the men’s or the women’s interviews. Results from this operation could readily be used to confirm gender stereotypes, as Table 1 illustrates. While the definition and interpretation of this version of ‘keyness’ are not without their critics (e.g. Moon 2007), it is clear that this variable-based analysis does tell us something about the data. It is a fairly blunt instrument, however, and is very prone to the methodological implication—even if this is not explicit—that there is some kind of causal link between the attribute of being male (or female) and using, in the context of telling your life story, more words from the domains of work and sport (or of the home and relationships). As previous researchers into language and gender have suggested, there may indeed be correlations between gender and conversational topic (e.g. Kipers 1987; Bischoping 1993; Tannen 1994)—and hence contrasting frequencies of words from particular domains. However, the realist social theory with which this analysis is concerned would seek to do something rather different from establishing such tendencies. For an explanation, rather than

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

It may well have a more important role in relation to the interpretation of the products of experiments in domains where experiments are useful, but when we deal with the products of surveys, with accounts of the world as becoming—that is, as a set of nested complex evolutionary systems, as inherently dynamic—then mundane exploration is as far as it goes.

A. SEALEY

223

Table 1: Selected lexical items from the top of the lists of words defined by WordSmith Tools as key in a comparison of the male and female interviews Keyness

Females

Keyness

wife period trade football film people boats number british country city company

200.54 170.11 139.06 128.04 124.09 119.78 119.34 114.89 106.03 98.606 93.962 89.773

husband children mom lovely baby home nursery know went mother pregnant sister

690.67 393.69 247.78 209.22 183.93 154.47 146.31 122.61 114.48 112.4 108.65 105.03

a description, of the variability in these different speakers’ accounts of their lives, a different methodology is called for. This is because: In the realist frame of reference we do not see causes as single factors whose presence inevitably generates an effect and whose absence means that the effect does not occur. Rather cause is a property of complex and contingent mechanisms in reality and such mechanisms, moreover, are not universal but only relatively permanent—inherently local. (Byrne 2002a: 105) Byrne’s research orientation derives from the sociological realism stimulated by Bhaskar and developed in various ways by Archer, Layder, Pawson, Sayer, and others. These writers ‘reject the conception of society . . . as [a] closed system, arguing instead that reality is a structured open system’ (Downward and Mearman 2007: 87). There is not the space here to elucidate the theoretical underpinnings of this approach (for further discussion see Sealey and Carter 2004; Sealey 2007), but an example of the method in action is instructive.

COMPLEX CAUSALITY AND SOCIAL BEHAVIOUR Realist explanations tend to be ‘theory driven approaches’ which seek to understand the mechanisms which lead to differential outcomes—why, for example, social or educational policy interventions turn out differently, appearing to ‘work’ for some people and not others, and in some circumstances and not others.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Males

224 PROBABILITIES AND SURPRISES

An excellent summary and explanation of this approach is found in Pawson (2002a). In public policy, meta-analyses of previous studies are often appealed to for answers to the question ‘what works’, but this is reformulated by Pawson as ‘what works for whom in what circumstances?’. He advocates a ‘realist synthesis’ perspective, according to which: . . . it is not ‘programmes’ that work: rather it is the underlying reasons or resources that they offer subjects that generate change . . .. Whether the choices or capacities on offer in an initiative are acted upon depends on the nature of their subjects and the circumstances of the initiative . . .. (Pawson 2002a: 342)

. . . adopts a ‘generative’ understanding of causation. What this tries to break is the lazy linguistic habit of basing evaluation on the question of whether ‘programmes work’. In fact, it is not programmes that work but the resources they offer to enable their subjects to make them work. This process of how subjects interpret the intervention stratagem is known as the programme ‘mechanism’ and it is the pivot around which realist evaluation revolves. (Pawson 2002a: 342) Studies of the social world from this kind of perspective offer an explanatory power that those based on variables do not. Applied linguists will be very familiar with studies designed to identify whether particular language teaching strategies ‘work’—and with the range of conflicting findings that they generate. The realist approach, by contrast, does not begin with social categories decided a priori—gender (or sex), or ethnicity, or age-group or social class. Instead, it assumes that social phenomena are characterized by processes and relations, and therefore ways are sought to investigate and describe these and their effects. If social phenomena are complex systems, then they: . . . are not made up of pre-existing variables, although we can properly describe them through the measurement of variate traces. In realist terms, the traces are actual things in the world

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Pawson describes the pattern in policy-making whereby a particular approach is tried in some area of public policy—such as the ‘naming and shaming’ of offenders, for example—and is then adopted in a whole range of other areas, as though the policy, the naming, and shaming, is causally effective (see also Pawson 2002b). This fallacious reasoning can lead to contradictory outcomes: for example, the car manufacturer whose products are publicized for their poor safety or security records is quick to improve their performance (‘successful’ naming and shaming), whereas the protesters who refuse to pay the community charge (a controversial local tax) find welcome celebrity in publicity surrounding their court appearances and are not persuaded to change their behaviour (‘unsuccessful’ naming and shaming). Pawson’s approach provides a persuasive explanation of the discrepancies, because it:

A. SEALEY

225

that are the products of the generative real system, and the interior working of the system is not reducible to elements existing separate and analyzable outside the system. (Byrne 2001: 64)

RESEARCHING COMPLEX PROCESSES—SOCIAL AND LINGUISTIC Readers who are familiar with corpus linguistics may identify some echoes here in the challenge to traditional descriptions of language that its methods and findings have generated. Language corpora (large collections of authentic writing or transcribed utterances, electronically stored) are analysed using dedicated software that reveals patterns not usually available from intuition, introspection or even text-by-text discourse analysis. Reliance on intuition rather than on the large quantities of empirical evidence available in the corpus, says Sinclair, meant that the pre-corpus situation in linguistics ‘was similar to that of the physical sciences some 250 years ago’ (1991: 1). In social science, maintains Byrne (2002a: 63), ‘our science for 300 years has been a science of analytical reduction to the simple. Now we can address the complex’, as computers ‘enhanc[e] our capacity to explore’—as they do with language. As realist theorists do not expect to observe directly the generative mechanisms responsible for differential social behaviour, so corpus linguists are familiar with the ‘tools of indirect observation’ (Sinclair 2004: 189) in their analytic software.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

It is appropriate at this point to summarize the methodological implications of the social realist theory within which the present study is situated. From Pawson’s work (which is predominantly concerned with the evaluation of interventions in social policy), an obvious priority is the identification of contexts and mechanisms to explain stability and variation in outcomes. The interpretations of experience by social actors may be consistent or varied, and patterns in both experience and interpretation are sought in the research process. According to Byrne (2001), each of the following is important: (i) exploration, ‘which involves both descriptive measurement of variate traces of complex systems and examination of the patterns generated by those measurements,’ including ‘the exploration of qualitative materials presented as texts or in other documentary forms’; (ii) classification, including sorting things into kinds, ‘using, inter alia, numerical taxonomy procedures’, as well as ‘the identification, however temporary, of what constitutes meaningful boundaries’; (iii) interpretation—both of measures and of ‘ ‘‘natural language’’ descriptions of qualitative form’; (in relation to this point, Byrne stresses that he is not referring to ‘post-modern eclecticism, but rather to the originary conception of hermeneutics in which there is a search for meaning as truth’; (iv) ordering, so that things are ‘sorted and positioned along a dimension of time’.

226 PROBABILITIES AND SURPRISES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The social researchers on whose work I am drawing are sympathetic to many different research methods, and several stress the advantages of moving between statistical analysis and local interpretation. For social scientists who want to know the reasons for patterned behaviours, ‘there is no logical difference between the work of an interpretive researcher conducting detailed observations of a social setting and a large scale national survey’ (Williams 2002: 128). Byrne (2002b: 68–69) sees a particular role for information technology here, ‘as a cybernetic extension of human cognitive capacity which enables us to get to grips with the big pictures of macroscopic social change through the management and interpretation of both quantitative and qualitative data flows’. Similarly, corpus linguists often emphasize the importance of moving between the corpus and the instance, combining automated searches and interpretive analysis (e.g. Upton and Connor 2001; Hunston 2002; Carter 2004: 219–221). Where conventional accounts of language systems separate lexis and syntax (‘a highly generalized formal syntax, with slots into which fall neat lists of words’ (Sinclair 1991: 108)), corpus analysis, which allows the researcher to review a lot of linguistic evidence at once, challenges traditional categories including word classes (or ‘parts of speech’). It also identifies trends and probabilities, which, according to Stubbs (2001) are attributable to the constraints that derive from both linguistic and social expectations. While not deterministic, these constraints ‘. . . mean that, although we are in principle free to say whatever we want, in practice what we say is constrained in many ways. The main evidence for these constraints comes from observations of what is frequently said, and this can be observed, with computational help, in large text collections’ (Stubbs 2001: 19). de Beaugrande (1997: 130) posits something similar—‘a convergence of data making some meanings or understandings much more probable than others’, and an explanation for ‘the rich global complexity of real language data’ as generated by ‘the interaction of multiple local constraints that are essentially simple’ (de Beaugrande 1999: 131). Complexity, dynamic processes, and relational units of analysis are critical to understanding both linguistic and social systems—yet both demonstrate stability: ‘as insights from corpus linguistics show, the stabilities that speakers employ are diverse—words, phrases, idioms, metaphors, noncanonical collocations, grammar structures—a much more complex and diverse set of language-using patterns than the ‘core grammar’ of formal approaches’ (Larsen-Freeman and Cameron 2008: 99). I am suggesting, then, that there are some parallels to be drawn between developments in researching social and linguistic processes, respectively, with reference to realist social theory and methodologies that are consistent with it. This is, of course, not to be blind to the dangers of ‘illegitimate crossdisciplinarity’, and indeed, as I have noted elsewhere (Sealey 2007), since different kinds of things have different properties and propensities, it is important to be mindful of those characteristics of language that are distinctive from those of social actors and of social structures. Nevertheless, the enterprise

A. SEALEY

227

described here aims to explore how far the analysis of a corpus of transcribed oral histories can be enhanced by the application of the integrative methodological approaches advocated by social realism within a corpus linguistic framework.

FROM CORPUS TO CASE—AND BACK AGAIN

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

One of the strategies proposed by Byrne is to start from the case, and to find ‘ways of sorting cases into categories’, looking for ‘category sets which emerge from the exploration of our data’ (Byrne 2002a: 100). With corpus data, such an approach can be achieved in various ways, and indeed Byrne’s injunction is consistent with Sinclair’s plea to ‘trust the text,’ including allowing computers to ‘. . . show us things that we may not already know and even things that shake our faith quite a bit in established models, and which may cause us to revise our ideas very substantially’ (Sinclair 2004: 23). Sinclair’s corpus-driven approach to language analysis mistrusts the premature tagging of corpus texts because of the danger of imposing prior models (or, as Byrne might say, ‘category sets’) and obscuring ‘the clarity of the categories in the data’ (Sinclair 2004: 191). The demographic categories associated with the corpus in the present study should therefore be treated with some caution—although it has to be noted that it is challenging to find useful groupings that maintain an appropriate balance between the distortion of over generalization and the unwieldy and unhelpful method of treating each case as unique. (A similar problem faces lexico-grammarians trying to be sensitive to specific patterning while simultaneously identifying groups with similar characteristics (Hunston and Francis 2000).) The longer term goal of the work under discussion is to combine corpus analysis and realist-derived sociological analysis through software adapted to accommodate both kinds of data. In the meantime, the approach is exemplified on a more modest scale using currently available tools. As a first step to a case-oriented methodology, the demographic meta-data about the 144 interviewees is entered into a table, with each row representing a case. This is consistent with Ragin’s (1987) advocacy of the comparative method to ‘compare cases with different combinations holistically’ (p. 101). ‘The case-oriented approach,’ he maintains, ‘. . . allows investigators to comprehend diversity and address causal complexity’ (p. 168). The next stage is to make use of established corpus analytic techniques to identify potentially fruitful areas of inquiry. To illustrate this, let us return to the contrasting keywords between male and female interviewees. A lexical word identified in this search was lovely, used four times as often by the female speakers as by the men, at 441 occurrences to the males’ 110. It is customary for quantitative researchers to stress that such findings indicate an association and not an explanation, and yet the inference that the attribute of being female is in some way causally linked with this language practice is not unreasonable. However, if being female was a direct ‘cause’ of the frequent

228 PROBABILITIES AND SURPRISES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

use of this word, then no men would use it often, and all the women would. This is patently absurd, but we can look somewhat differently at this correlation. Counter-posed to the implicit determinism of the variables-based approach are the more interpretive research traditions. Instead of inferring that possession of the attribute [variable] female leads to an increased use of items such as lovely, some analysts would stress the performative dimension of gender through language (e.g. Ochs 1992; Meyerhoff 1996; Holmes and Meyerhoff 1999; Weatherall 2002). Holmes (1997: 203), for example, suggests that attention should be on ‘the linguistic realizations of gender’, through an examination of ‘the way individuals express or construct their gender identities in specific interactions in particular social contexts’. More ethnographic, qualitative studies, such as those advocated by Holmes, focus on ‘how people use language to create, construct, and reinforce particular social identities’ (Holmes 1997: 204)—and women’s frequent use of positive evaluators such as lovely may well index femininity (cf. Lakoff 1975). However, such approaches are rarely concerned with the probabilities identified in quantitative analysis, and, while they do shed light on the local development of communities of (more or less gendered) practices, their accounts of speakers resisting the reproduction of gendered patterns of language behaviour tend not to propose causal explanations. Returning to the individual cases in my corpus, we can look more closely at the counter-examples, the cases that seem to buck the trend—rather than regarding them as ‘outliers’, as conventional statistical procedures would require. These ‘outliers’ include, from the 16 speakers who use lovely in their interviews most frequently (10 times or more), the two who are men. From a case perspective, it transpires that there are similarities across several dimensions. Both were born in Birmingham in the 1920s; neither continued their education beyond secondary school; both are fathers to two or more children and both are married. Further exploration could establish whether there are additional features, linguistic or demographic, which would identify reasons to see these speakers as belonging to a sub-group linked in meaningful ways; alternatively, it may be that choosing lovely to describe elements of their experience, when most men do not do so in these interviews, is a weak link, not reinforced by any others. Looking at the same initial findings from the other direction, so to speak, it transpires that a significant number—59—of the 144 speakers fail to use the word lovely in their interviews at all. Are these non-users all men? The majority—44—are, but the remaining 15 are women. From a case-based perspective, we can investigate whether these speakers have anything in common, looking for both linguistic and demographic patterns. One possibility that suggests itself, of course, is that these particular women have not experienced anything that could be described as ‘lovely’. Perhaps (their accounts of) their biographies are distinctively negative. To investigate this possibility, I used WMatrix to compare each of these 15 texts

A. SEALEY

229

(a)else, kind of thing. We had fun when we moved here obviously so all that. I just go out for a laugh with my friends and then just g oose to go somewhere where you enjoy yourself rather than the fact of a day out and that was really good fun because it was really relaxed ried and having kids and being happy with it because that ’s not me s quite good, but it was just fun but then you got sick of it in ple that I know so it ’s quite fun in that respect but as the job we have kind of got the same sense of humour, so we hang round mostly at you want to do, so you can enjoy yourself a lot more this time in so it was quite good. It was enjoyable because it’s in the country an

(b)the drama department is a for a month which was out walking and it was was Amsterdam which was it ’s always just been

brilliant at my school and if brilliant and we got to travel brilliant and we flew &; I think brilliant and I went to see Anne brilliant , there ’s never been

(c) to accept it. And my mom who’s fab was like I don’t care e her and she loves you. She was fabulous but found it modation we wanted, that was the best route, doing that. award and I won which was really fabulous, probably one probably one of the best moments of my life which was been to get there really. It was fab .

Figure 2: (a) Selected concordance lines for semantic domain of ‘happy’ in one interviewee’s transcript. (b) Selected concordance lines for semantic domain of ‘evaluation_good’ in another interviewee’s transcript. (c) Selected concordance lines for semantic domain of ‘evaluation_good’ in a third interviewee’s transcript

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

with the corpus as a whole, using the semantic domains tool described above. This allows the user to see in a concordance line view each of the items assigned to the semantic domains identified by the tagger. From this, it is apparent that some of these speakers seem to be very positive in their narratives, using items classified in the ‘evaluation_good’ category and in the ‘happy’ category significantly more frequently than is found in the corpus as a whole. Some examples from the concordance lines for these data are given below (see Figure 2). This suggests that the interplay between these women’s oral histories and their lexical choices is more subtle than an absence of lovely simply denoting an absence of positive evaluation in general. From the case-based perspective, the table of metadata can be reviewed manually, because the number of cases is so small, although it would be desirable to use appropriate software for a larger-scale analysis. Of all the attributes recorded in the database, just one emerges as shared among all but two of this particular group of women: their classification by ‘marital status’ is single or widowed. The remaining two have been classified as ‘other’: one is a gay woman who describes in her interview the difficulties she has faced because of her sexuality, and

230 PROBABILITIES AND SURPRISES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

another, who is in a stable heterosexual relationship, reflects explicitly on her resistance to getting married and to conforming to conventional gender roles. Thus the combination of these two approaches, a case-based consideration of the interviewees’ attributes, and a corpus-based linguistic analysis, generates a potential sub-group who share the classification ‘female’, who significantly under-use the item lovely in their interviews, relative to other females in the corpus, and who are not currently in conventional marital partnerships. A further iteration of the approach makes use of the keyword facility to compare the sub-corpus of the interviews of these women with the corpus as a whole. This makes possible, for example, the identification of other items, in addition to lovely, that are frequent in the accounts of the other female speakers, but much less so in these interviews. Items that WordSmith identifies as ‘negatively key’ (that is, significantly under-used by this sub-group in comparison with the sub-corpus of the other female interviewees) include child, mother, wonderful, husband, children—and also him, his and he. Very near the top of the list of items they use more frequently than the other female speakers is don’t, and a concordance of I don’t is interesting. There are 427 occurrences in the interview transcripts of these 15 women, many of which frequently occur in strings such as ‘I don’t know’. However, the most frequent four-word string (27 occurrences) is ‘I don’t want to’, while there are 29 occurrences of the three-word ‘I don’t like’. All speakers use one of these strings, 10 of them using only either ‘I don’t like’ or ‘I don’t want to’, and the other five using both. Looking in detail at the concordance lines for these strings, it becomes apparent that these particular women tell their life stories with repeated references to the differences between themselves and others, or between their priorities and the expectations or priorities of the wider society. To illustrate this, just one example has been extracted of these two expressions from each of the speakers in this group who uses them (see Figure 3). This finding receives further confirmation from the analysis of semantic domains facilitated by WMatrix. Domains identified as occurring significantly more frequently in the interviews of some of these 15 speakers include ‘not part of a group’ and ‘different’. Some examples of the concordance lines classified in this way are presented in Figure 4. Much more analysis would be needed to draw any definitive conclusions, but this case-based, corpus-informed analysis has begun to identify a subgroup of speakers who tell their life stories in a way that would not have been identified using any of these analytic approaches on its own. As women, they are somewhat unusual in failing to use in their interviews such positive evaluators as lovely and wonderful. None of them is currently in a conventional marital relationship, and they recognize and describe themselves as often different in their behaviour or outlook from the norms—though not exclusively the gender norms—presented by others.

A. SEALEY

231

(a) I don’t like the current English culture I don’t like housework very much. I don’t like ironing, I don’t like washing I’m very direct and I don’t like to hide because I’m not embarrassed about who I am. I’ll move away from Birmingham anyway, just because I don’t like the city. I like open spaces but I don’t like the sea Birmingham is a very, and I don’t like to use the word, multi-cultural environment. I don’t like that my disability has [been?] made the subject of competitive spirit I don’t like fighting or anything. I don’t mind drinking but I don’t like fighting, so that’s it. I thought I am not going back to this, so I thought, no, I don’t like this I don’t like to mix much, you know

(b)

Figure 3: (a) Concordance lines for ‘I don’t like’ from the interviews of ten different speakers. (b) Concordance lines for ‘I don’t want to’ from the interviews of ten different speakers In these brief extracts, I suggest, we gain some brief glimpses into what Archer (2000: 163) identifies as ‘one of the most important things to probe’, namely ‘how the self-conscious human being reflects upon his or her involuntary placement’. A realist, agency-based perspective has no problem accounting for this complex configuration of findings as an example of people making choices from the resources and options available to them— and in a patterned way that does not necessarily conform to the rather inflexible categories conventionally deployed in quantitative studies. It is more consistent with the ‘performance of gender’ perspective familiar from qualitative research, but allows for different levels of analytic purchase on the data.

CONCLUSION In this article I have sketched briefly the ways in which social researchers seek to account for probabilistic patterns with reference to realist social theory. This theory has several important features. It conceives of the social world as an open, complex, dynamic set of inter-related systems. Human behaviour is understood to be explicable not with reference to single causes that are effective in categorical ways, but with reference instead to nested and interacting sets of interests and circumstances, some of them involuntary and perhaps even unknown to the people affected by them, others the results of choices made with reference to what people perceive as in their interests. In this approach, social category labels are seen as neither discursive ephemera nor deterministic causal variables. Many of the categories routinely used in survey research, policy evaluations and monitoring practices are less stable than they may seem at first sight (Sealey and Carter 2001; Carter and Sealey in press). The social category used as an illustration here was that of ‘gender’. Once assumed to be an essential, deterministic attribute firmly linked to

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

I’ve never visited the country and I really think that I don’t want to because I’ve seen too much of a bad side So, when he came I said I don’t believe in unions so I don’t want to join. Well most of the others had joined fair enough, people do want to get married, but personally I don’t want to waste all that money on getting married I don’t know how you’re supposed to change that and I don’t want to change being a confident person I don’t want to be ignored, but I don’t want to be looked after either I was thinking but I don’t want to be like that, I want to be myself I don’t want to set my sights too high I didn’t want to do any of that, I just thought no, I don’t want to be anything in medical oh, oh no, oh no, I don’t want to go to work it just feels like you are not really needed, and so whatever I decide to do, I don’t want to be a university lecturer

232 PROBABILITIES AND SURPRISES

(a)felt like strangers in a way or outsiders because we didn’t realise t once, because I felt such an outsider because I didn’t listen to ulated this feeling of being an outsider because they obviously had a overwhelming sense of being an outsider and not wanting to be an I was much more solitary in that sense and it was in 1988 I felt like even more of an outsider because I had a different perspe

(c) I always had a sense that I was different from everyone else from asked me to ask you why you’re different to us. And I think that colour. I knew I was different and at home, our life at home wa communicated with her family was very different to the way that I’d ifferent, you know language was different and even then, I mean I but I knew things were different in the way that they acted within great deals of class differences and certain children doing better you get compared to the other and that might have given her a b university, that makes you different and you’re not the same person a and so I was quite horrified at other people’s personal habits, I thi r in’92, I did n’t realise how segregated the city was and did n’t

Figure 4: (a) Selected concordance lines for semantic domain of ‘not part of a group’ in one interviewee’s transcript. (b) Selected concordance lines for semantic domain of ‘not part of a group’ in a second interviewee’s transcript. (c) Selected concordance lines for semantic domain of ‘comparing_different’ in a third interviewee’s transcript

biological sex, gender later came to be understood as a product of socialization and acculturation, while current perspectives emphasize the notion of performed, ‘gendered’ identities, and of diversity in place of binary difference (Cameron 2005). The detailed life histories discussed above reveal diversity in relation not only to gender identities, but also religious affiliation, interpretations of ‘ethnicity’, people’s status as ‘parents’ (where the role may or may not be biological, legal, adoptive, temporary and so on) and many other social categories which are often used to assign an individual to membership of a group. Recognition of both the diversity and the commonalities that underlie such social categories raises a crucial question: to what extent are we ‘positioned’ by ‘external socio-cultural factors which . . . predispose us to various courses of action’ (Archer 2000: 12), and how much scope do individuals have to choose and fashion for themselves one or more identities? Archer (2000: 13) puts it this way. ‘Society enters into us, but we can reflect upon it, just as

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(b)really. Independence was the one thing I enjoyed at s ‘t think that was right, but I personally feel being at an all girls n all girls ‘ school I was more independent, more so than going to a I find that I was more independent than some people that I ‘d met w anything individually or personally but I believe there is something o get the experience. For me, personally, a lot of people who have e that in London, it ‘s a very lonely place . And the other thing Television and $pla and $pla as freelancer in camera production

A. SEALEY

233

ACKNOWLEDGEMENTS I am grateful to Sian Roberts, Helen Lloyd, and Malcolm Dick for access to, and information about, the Millennibrum interviews, as well as to the people who recorded them for posterity. I should also like to thank Pernilla Danielsson, Paul Rayson, and the Collaborative Research Network at the University of Birmingham which funded the post-editing of the transcripts, most ably carried out by Juliet Herring. Grateful acknowledgements are also due to Bob Carter, three anonymous referees and the editors of the journal for helpful comments on an earlier version of this article.

REFERENCES Archer, M. 2000. Being Human: The Problem of Agency. Cambridge University Press.

de Beaugrande, R. 1997. New Foundations for a Science of Text and Discourse: Cognition,

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

we reflect upon nature and upon practice. Without such referential reality there would be nothing substantive to reflect upon; but without our reflections we would have only a physical impact upon reality.’ Contemporary research that is consistent with this theoretical outlook makes use of various methods, including the ever-increasing processing capacity of computers. The methods used are iterative, applying different scales of analysis, as is consistent with a view of the social world as comprising different kinds of entities with different properties and powers that operate on different timescales. I have suggested that language and discourse, as indispensable elements of both social processes and the researching of them, can yield particular kinds of insight, and that there are parallels between corpus-driven linguistic analysis and research into social processes that, like corpus linguistics, is open to the data-driven identification of categories, patterns, and probabilities. With reference to the transcribed life histories of a particular collection of people, I have demonstrated how these approaches to research have the potential to be mutually informative. The illustrations provided are necessarily selective and tentative, and much more will need to be achieved before the full potential of the optimum software for bringing together linguistic and sociological analysis is likely to be realized. For example, the ‘variable’ of sex (or gender) was chosen as a convenient initial category to interrogate, and others from the demographic data would no doubt generate equivalent lines of inquiry. The ‘positive evaluators’ chosen as the starting point for the linguistic analysis could have been substituted by any number of alternative patterned linguistic features (current work in progress is looking at hedges and boosters, vague expressions, significantly frequent grammatical words and so on). And the tools of analysis, including calculators of ‘keywords’, will ideally be extended as this approach develops. Nevertheless, the potential identified for linguistic and social researchers to get the measure of their data in innovative ways is an exciting prospect.

234 PROBABILITIES AND SURPRISES

Holmes, J. and M. Meyerhoff. 1999. ‘The community of practice: theories and methodologies in language and gender research,’ Language in Society 28: 173–183. Hunston, S. 2002. Corpora in Applied Linguistics. Cambridge University Press. Hunston, S. and G. Francis. 2000. Pattern Grammar: A Corpus-Driven Approach to the Lexical Grammar of English. John Benjamins. Kipers, P. S. 1987. ‘Gender and topic,’ Language in Society 16: 543–557. Lakoff, R. 1975. Language and Woman’s Place. Harper and Row. Larsen-Freeman, D. and L. Cameron. 2008. Complex Systems and Applied Linguistics. Oxford University Press. Layder, D. 1997. Modern Social Theory: Key Debates and New Directions. UCL Press. Meyerhoff, M. 1996. ‘Dealing with gender identity as a sociolinguistic variable’ in V. L. Bergvall, J. M. Bing, and A. F. Freed (eds): Rethinking Language and Gender Research: Theory and Practice. Addison Wesley Longman. Moon, R. 2007. ‘Words, frequencies, and texts (particularly Conrad): A stratified approach,’ Journal of Literary Semantics 36: 1–34. Ochs, E. 1992. ‘Indexing gender’ in A. Duranti and C. Goodwin (eds): Rethinking Context: Language as an Interactive Phenomenon. Cambridge University Press. ONC (Office for National Statistics) http://www. ons.gov.uk/about-statistics/classifications/ current/SOC2000/about-soc2000/index.html. Accessed July 2008. Pavlenko, A. 2007. ‘Autobiographic narratives as data in applied linguistics,’ Applied Linguistics 28/2: 163–188. Pawson, R. 2002a. ‘Evidence-based policy: The promise of ‘realist synthesis’ ,’ Evaluation 8/3: 340–358. Pawson, R. 2002b. ‘Evidence and policy and naming and shaming,’ Policy Studies 23/3–4: 211–230. Ragin, C. C. 1987. The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies. University of California Press. Rayson, P. 2008a. Wmatrix: A Web-Based Corpus Processing Environment. Computing Department, Lancaster University. http://ucrel.lancs.ac.uk/ wmatrix/

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Communication, and the Freedom of Access to Knowledge and Society. Ablex. de Beaugrande, R. 1999. ‘Linguistics, sociolinguistics, and corpus linguistics: ideal language versus real language,’ Journal of Sociolinguistics 3/1: 128–139. Bischoping, K. 1993. ‘Gender differences in conversation topics, 1922–1990,’ Sex Roles 28/1–2: 1–18. Byrne, D. 2001. ‘What is complexity science? Thinking as a realist about measurement and cities and arguing for natural history,’ Emergence 3/1: 61–76. Byrne, D. 2002a. Interpreting Quantitative Data. Sage. Byrne, D. 2002b. ‘Platonic forehand versus Aristotelian smash—the use of computers as macroscopes in knowing the social world,’ International Journal of Social Research Methodology 5/1: 61–69. Cameron, D. 2005. ‘Language, gender and sexuality,’ Applied Linguistics 26/4: 482–502. Carter, B. and Sealey, A. in press. ‘Reflexivity, realism and the process of casing’ in D. Byrne and C. C. Ragin (eds): Handbook of Case-based Research Methods. Sage. Carter, R. 2004. Language and Creativity: The Art of Common Talk. Routledge. Craib, I. 1998. Experiencing Identity. Sage. Dick, M. (ed.) 2002. Millennibrum: Bringing Birmingham History to Life CD-ROM. Birmingham City Council. Downward, P. and A. Mearman. 2007. ‘Retroduction as mixed-methods triangulation in economic research: Reorienting economics into social science,’ Cambridge Journal of Economics 31: 77–99. Grice, H. P. 1975. ‘Logic and conversation’ in P. Cole and J. L. Morgan (eds): Syntax and Semantics; Vol 3: Speech Acts. Academic Press. Halliday, M. A. K. 1989. ‘Towards probabilistic interpretations’ in E. Ventola (ed.): Functional and Systemic Linguistics: Approaches and Uses. Mouton de Gruyter. Halliday, M. A. K. 1991. ‘Corpus studies and probabilistic grammar’ in K. Aijmer and B. Altenberg (eds): English Corpus Linguistics. Longman. Holmes, J. 1997. ‘Women, language and identity,’ Journal of Sociolinguistics 1/2: 195–223.

A. SEALEY

Sealey, A. and B. Carter. 2004. Applied Linguistics as Social Science. Continuum. Sinclair, J. 1991. Corpus, Concordance, Collocation. Oxford University Press. Sinclair, J. M. 2004. Trust the Text: Language, Corpus and Discourse. Routledge. Sperber, D. and D. Wilson. 1986. Relevance. Blackwell. Stubbs, M. 1996. Text and Corpus Analysis. Blackwell. Stubbs, M. 2001. ‘Texts, corpora and problems of interpretation,’ Applied Linguistics 22/2: 149–172. Tannen, D. 1994. Gender and Discourse. Oxford University Press. Uprichard, E. and D. Byrne. 2006. ‘Representing complex places: A narrative approach,’ Environment and Planning A 38: 665–676. Upton, T. A. and U. Connor. 2001. ‘Using computerized corpus analysis to investigate the textlinguistic discourse moves of a genre,’ English for Specific Purposes 20/4: 313–329. Weatherall, A. 2002. Gender, Language and Discourse. Routledge. Williams, M. 2002. ‘Generalization in interpretive research’ in T. May (ed.): Qualitative Research in Action. Sage. Wortham, S. 2000. ‘Interactional positioning and narrative self-construction,’ Narrative Inquiry 10: 157–184.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Rayson, P. 2008b. ‘From key words to key semantic domains,’ International Journal of Corpus Linguistics 13/4: 519–550. Rayson, P., D. Archer, S. L. Piao and T. McEnery. 2004. The UCREL semantic analysis system. In Proceedings of the Workshop on Beyond Named Entity Recognition Semantic Labelling for NLP Tasks in association with 4th International Conference on Language Resources and Evaluation (LREC 2004), 25th May 2004, European Language Resources Association, pp. 7–12. Rumelhart, D. E. 1975. ‘Notes on a schema for stories’ in D. G. Bobrow and A. M. Collins (eds): Representation and Understanding: Studies in Cognitive Science. Academic Free Press, pp. 211–236. Rumelhart, D. E. 1984. ‘Schemata and the cognitive system’ in R. S. Wyer and T. K. Srull (eds): Handbook of Social Cognition. Lawrence Erlbaum, pp. 161–88. Scott, M. 2004. WordSmith Tools Version 4. Oxford University Press. Sealey, A. 2007. ‘Linguistic ethnography in realist perspective,’ Journal of Sociolinguistics 11/5: 641–660. Sealey, A. and B. Carter. 2001. ‘Social categories and sociolinguistics: Applying a realist approach,’ International Journal of the Sociology of Language 152: 1–19.

235

Applied Linguistics: 31/2: 236–259 ß Oxford University Press 2009 doi:10.1093/applin/amp024 Advance Access published on 4 June 2009

Lexical Diversity in Writing and Speaking Task Performances GUOXING YU University of Bristol, UK

INTRODUCTION Lexical diversity has been considered an illuminative predictor of learners’ general language proficiency (e.g. Zareva et al. 2005) and an essential indicator of the quality of their writing (e.g. Laufer and Nation 1995) and speaking (e.g. O’Loughlin 1995; Jarvis 2002; Malvern and Richards 2002) performances. Such a positive relationship is also claimed explicitly in the rating scales of major international language tests, as well as in the development of automated evaluation systems. For example, IELTS writing and speaking responses

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

In the rating scales of major international language tests, as well as in automated evaluation systems (e.g. e-rater), a positive relationship is often claimed between lexical diversity, holistic quality of written or spoken discourses, and language proficiency of candidates. This article reports a posteriori validation study that analysed a sample of the archived data of an international language test to examine empirically to what extent such relationships exist. It is also noted that previous studies on lexical diversity in the field of applied linguistics have focused exclusively on either written or spoken discourses, no study to date has compared lexical diversity of spoken and written discourses produced by the same participants. Therefore, the second aim of this article is to understand the differences in lexical diversity between writing and speaking task performances, and to what extent the topics of the writing prompts may affect lexical diversity of written discourses. Using D as a measure of lexical diversity (Malvern and Richards 1997, 2002; Malvern et al. 2004), it was found that D had a statistically significant and positive correlation with the overall quality ratings of both writing and speaking performances as well as the candidates’ general language proficiency. Nevertheless, the significant relationships were not borne out across the subgroups of the sample in terms of gender, first language background, purpose of taking the test and topics of the writing prompts. The different writing topics also had significant effects on lexical diversity—especially the topics that candidates were highly familiar with—even after controlling for writing ability and overall language proficiency. The lexical diversity of candidates’ writing and speaking performances were approximately at the same level; further, D was found to be a better predictor of speaking than writing performance. The implications of these findings are discussed with specific reference to the use of lexical diversity measures to inform language test validation and the development of lexical diversity parameters in automated evaluation systems.

G. YU

237

RELATED LITERATURE This section provides a short account of different methods of measuring lexical diversity, with a specific focus on D (Malvern and Richards 1997, 2002; Malvern et al. 2004). It then reviews in detail several empirical studies on the relationships between lexical diversity and quality of written and spoken discourses.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

are rated according to their ‘lexical resource’ which refers to ‘the range of vocabulary the candidate has used’ (IELTS Handbook 2007). One of the TOEFL iBT criteria to score compositions is ‘appropriate and precise use of grammar and vocabulary’ (TOEFL iBT Scores 2005). According to the rating scales of Michigan English Language Assessment Battery (MELAB Technical Manual 2003), candidates need to demonstrate ‘a wide range of appropriately used vocabulary’ to achieve a high score for compositions, and similarly an ‘excellent speaker’ should demonstrate his or her use of ‘idiomatic, general, and specific vocabulary range is extensive’; and for a ‘marginal speaker’, his or her ‘vocabulary is limited’. Lexical diversity has also been employed as one of the most important parameters in automated writing (e.g. Chodorow and Burstein 2004) and speaking (e.g. Zechner et al. 2007) evaluation systems. It seems that lexical diversity has been widely assumed as an important quality indicator of test performance. However, the precise nature of such relationships remains largely asserted and sometimes elusive. Given this widespread assumption and the high stakes and impact of international language tests, posteriori validation studies are needed to establish empirically to what extent such explicitly claimed links exist in the actual test performance data. Furthermore, several other questions remain unanswered such as: to what extent are candidates’ lexical diversity in writing and speaking task performances related, and to what extent is lexical diversity of written or spoken discourses attributable to candidates’ overall language proficiency, and, further, what task characteristics may affect lexical diversity of performance and how. This article aims to address these questions by analysing a representative sample of the archived data of an international language test. However, a couple of limitations of this research should be acknowledged at the outset. Firstly, the quality of a discourse, written or spoken, is defined and shaped by various linguistic features (e.g. handwriting quality; pronunciation and fluency in speaking) other than lexical diversity alone. Secondly, non-linguistic factors such as candidates’ anxiety and stress (Ho¨weler 1972), anticipation of being evaluated (e.g. Jarvis 2002), educational attainment (Sankoff and Lessard 1975), and cognitive development stage (e.g. Dura´n et al. 2004) could affect lexical diversity (see Bradac et al. 1979 for a review), and therefore are equally important for understanding lexical diversity of test performance and the validity of the rating scales in relation to the use of vocabulary.

238 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

Defining lexical diversity

Measuring lexical diversity As with the nomenclature confusions, the methods for quantifying lexical diversity also seem to be both ‘prolific’ and problematic, not only within a discipline but across disciplines such as stylometry, psychology, speech pathology, forensic linguistics, corpus linguistics, and applied educational linguistics (e.g. language assessment and second language acquisition). Measuring lexical diversity is at least as complex as the concept itself and is ‘not a straightforward task’ (Dura´n et al. 2004: 220). Since Yule’s (1944) seminal work (see Laufer and Nation 1995; Malvern and Richards 1997; Vermeer 2000; Daller et al. 2003; Malvern et al. 2004 for an overview and critiques), various attempts have been made to develop conceptually sound and mathematically simple lexical diversity measures in many languagerelated research areas (e.g. Sichel 1986; Tweedie and Baayen 1998; Wimmer and Altmann 1999; Hoover 2003; Panas and Yannacopoulos 2004).

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

What, then, is lexical diversity? In the literature, there seems to be both nomenclature diffusions and confusions. One such confusion arises from the fact that this term has been used to refer to language abilities of producers (writers or speakers) as well as the quality of products (written or spoken). Although the producers and their products are apparently related, the lexical diversity of a product is only one static manifestation of the producer’s lexical diversity which may well be dynamic in nature.1 Further terminological problems arise from the fact that various similar terms (see Malvern et al. 2004: 3–15) such as lexical richness (e.g. Singh 2001; Daller et al. 2003), lexical density (e.g. O’Loughlin 1995), lexical sophistication or rareness (see Read 2000: 203), lexical variation or variety (e.g. Hyltenstam 1988; Granger and Wynne 2000), lexical individuality or originality, lexical complexity or simplicity, vocabulary diversity (e.g. Ciani 1976; Johnson 1979), lexical range and balance (Crystal 1982), vocabulary richness (e.g. Ha¨rnqvist et al. 2003; Sokolova et al. 2006), and vocabulary density are frequently used interchangeably with lexical diversity (e.g. Malvern and Richards 2002), but sometimes exclusively to each other, and sometimes hierarchically (e.g. Read (2000) defines lexical diversity, sophistication and variation as different aspects of lexical richness). Further complications arise when the same term was conceptualized and quantified differently in different studies. Indeed, different conceptualizations and quantifications of ‘lexical diversity’ make it difficult to compare and synthesize the findings of those studies although they used the same term—‘lexical diversity’. In this research, in line with the rating scales of the aforementioned international language tests, lexical diversity is loosely defined as one aspect of the quality of a product—as ‘something to do with the range of vocabulary displayed’ in written and spoken discourses (Dura´n et al. 2004: 220).

G. YU

239

D is the single parameter of a mathematical function that models the falling TTR curve. It is beyond the scope of this article to discuss D mathematically (see Chapter 3 of Malvern et al. (2004) for its rationale and mathematical derivations). In essence, the higher the D, the greater the diversity of a text. A computer program called vocd in CLAN (MacWhinney 2000) provides a standardized procedure for measuring D (see Malvern et al. 2004). The key methodological advantage of D over other measures is that D is not a function of sample size,2 thus allowing valid comparisons between varying quantities of linguistic data (see Malvern and Richards 2002: 91; Malvern et al. 2004). Jarvis (2002) found that D provided accurate typetoken curve fitting models of lexical diversity of short texts. He thus concluded that D seems to ‘have put us on the right track’ (p. 82) to measure lexical diversity (but see McCarthy and Jarvis 2007). Data from various sources (from children to adults, from L2 learners to academic writers, from normally developing children to those with specific language impairment) confirmed that D was capable of demonstrating the developmental trends observed (Dura´n et al. 2004). Studies in language impairment (Owen and Leonard 2002; Silverman and Ratner 2002; Wright et al. 2003; Klee et al. 2004) have also extensively confirmed methodological advantages of D.

Lexical diversity and linguistic factors in writing and speaking task performances This section reviews several empirical studies in applied educational linguistics that have conceptualized, either implicitly or explicitly, lexical diversity as an end-product, as a quality indicator of writing and speaking performance. It should be noted that due to the disparities in the conceptualizations and

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

TTR (type-token ratio) was once the most widely used across disciplines, however, TTR is notorious for being sensitive to sample size (i.e. TTR curve systematically falls with the increase of text length) and renders itself incapable of assessing lexical diversity in a satisfactory way if sample size varies a great deal (see Arnaud 1984; Richards 1987). Other measures, pffiffiffiffi through some algebraic transformations of TTR, e.g. Root TTR (T/ N ) by pffiffiffiffiffiffi ffi Guiraud (1954, 1960), corrected TTR (T/ 2N ) by Carroll (1964), Log TTR (log T/log N) by Herdan (1960) were developed, attempting to reduce the effects of sample size. However, Vermeer (2000: 76) found that ‘neither the validity nor the reliability of the various measures of lexical richness in direct data are satisfactory’. More recently, another approach has been taken by Malvern and colleagues which results in the following equation containing the parameter D (Malvern et al. 2004: 47): " #  D D 1=2 1þ2 1 TTR ¼ N N

240 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

quantifications of lexical diversity in these studies, their findings are sometimes not easily comparable. Linnarud (1986) found clear differences in the use of vocabulary between the compositions written by 17-year-old Swedish learners of English, who had been learning it for 9 years, and those by English native speakers of the same age. The Swedish learners lacked lexical variation and showed much less lexical originality than the English native speakers. In addition, there was a large difference between the number of individual words most frequently used by English native speakers and Swedish learners. The difference in nouns was the smallest, and the largest differences were in the use of adjectives and adverbs. In compositions written by EFL learners (intermediate to highintermediate level of proficiency), Engber (1995) found substantial correlations between lexical diversity and holistic quality ratings of the compositions. Vermeer (2000), Daller et al. (2003), and O’Loughlin (1995) examined lexical diversity of spoken discourses. Vermeer (2000) compared a series of measures such as number of types, TTR, Guiraud, Herdan and Uber indices applied to spontaneous speech—individual face-to-face interviews between himself and children (aged 4–7) learning Dutch as an L1 or L2. The correlations between lexical diversity of the spontaneous speech and learners’ performances in receptive and productive vocabulary tests demonstrated that although there seemed to be ‘a clear influence of number of tokens and number of types’ (p. 77), neither of these measures showed concurrent validity with both the receptive and productive vocabulary tests, particularly at learners’ later stage of vocabulary acquisition (i.e. vocabulary size greater than 3000 words). In other words, his research demonstrated that the relationship between lexical diversity of spontaneous speech and learners’ vocabulary knowledge was not clear. However, as I have argued earlier, this may well be due to the use of different lexical diversity measures. It remains to be seen whether D, unavailable then, might present a different picture of the correspondence between lexical diversity measures and the vocabulary tests. In fact, the use of D in some studies on language impairment (e.g. Owen and Leonard 2002; Silverman and Ratner 2002; Wright et al. 2003; Klee et al. 2004) have demonstrated that D was a better indicator of language developmental trend than other traditional lexical diversity measures. Similar to Vermeer (2000), Daller et al. (2003) applied TTR and Guiraud, combined with Laufer and Nation’s (1995) notion of different levels of lexical frequency and productivity by distinguishing basic and advanced lexical items, on spontaneous oral productions of Turkish-German bilinguals in higher education institutions in Germany and Turkey. Significant correlations were found between participants’ language proficiency and the lexical diversity based on advanced lexical items. O’Loughlin (1995) found ‘lexical density’—specifically defined as a measure of the relationship between grammatical items and high- and low-frequency lexical items of oral performance—was significantly affected by test formats (live or tape-recorded) and task types (e.g. describing a familiar setting and

G. YU

241

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

role-play) as well as the interactions of these test method facets. Between live and tape-recorded role play tasks was the most salient difference in ‘lexical density’, which O’Loughlin attributed to the fact that the interviewer and candidates exchanged information in the live task, while in the tape-recorded task, candidates were simply asked to leave a telephone answer-machine message. He suggested that ‘the higher the level of interaction the lower the degree of lexical density’ (p. 234), and that ‘the degree of interactiveness, rather than test format, emerges as perhaps the single most important determinant of candidate output’ (p. 236). Like Vermeer (2000) and Daller et al. (2003), Jarvis (2002) tested various lexical diversity measures on short written narratives by young EFL learners and English native speakers. The methodological advantages of D over other measures were again confirmed in this study. In this study, he used base forms of lexical items to measure D. It was found that the narratives by EFL learners with more years of English learning experience tended to have higher D. However, this developmental trend was less clear when using only the content words (i.e. with function words removed) in measuring D than when using all the words in narratives. Younger native speakers produced narratives of higher D than older native speakers. Nevertheless, significant difference between EFL learners and native speakers was noted, always with native speakers having a higher D. The follow-up tests, however, indicated ‘significant differences were found only between some groups of learners . . . and some groups of native speakers’ (p. 74). Jarvis suggested this result implied that ‘L1 background may not be a reliable predictor’ of lexical diversity. Overall, lexical diversity was found to have consistently significant, albeit moderate, correlations with the holistic quality ratings of the narratives. However, these statistically significant correlations held true only for the Finland-Swedish participants. For Finnish and Americans, the correlations were low and insignificant. In addition, Jarvis observed that written narratives of the highest mean D tended to have the lowest, even negative, correlation between D and holistic quality ratings. These studies together demonstrate that different conceptualizations would quite naturally lead to different quantifications of lexical diversity, and that different measures may well produce very different, sometimes even conflicting results. It also seems that the findings of these studies may be task specific—not only in terms of what types of elicitations were used (planned writing or spontaneous speech) but also how the tasks were initiated (live or tape-recorded) and managed (e.g. whether and what accommodation are provided by the interviewers). The relationship between lexical diversity and overall quality rating of a discourse may be also participantspecific. It is important to note that only Engber’s (1995) and O’Loughlin’s (1995) studies had participants from diverse L1 backgrounds (although these two studies did not examine whether L1 affected lexical diversity); the rest of the studies had focused on European learners of English or European languages per se. Jarvis’ (2002) implication that L1 may not be a reliable predictor

242 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

of lexical diversity could well be attributable to the closer ties between English and European languages. The possible effect on lexical diversity of language distance (see Elder and Davies 1998) between learners’ L1 (e.g. Chinese) and English might not be negligible. It is equally important to note that the studies reviewed here had focused on lexical diversity of either written or spoken texts; but none of them had investigated lexical diversity of both written and spoken texts by the same participants. Indeed to the best of my knowledge, no research to date has been conducted in the field of language testing that compares lexical diversity of spoken and written discourses produced by the same learners of English.

Research questions This research analysed a sample of MELAB archived data (January 2004 to November 2005) to investigate the lexical diversity of their writing and speaking performance. Specifically, five research questions (RQ) addressed whether and to what extent the lexical diversity of: 1 2 3 4

the written discourses and the scores assigned to them are related? the written discourses is affected by the writing topics? the spoken discourses and the scores assigned to them are related? the written and the spoken discourses are related to the candidates’ overall language proficiency? 5 the written and the spoken discourses are related? In addition to the main focus above, candidates’ characteristics including their gender, purpose for taking the test and L1 background are also taken into consideration in the analyses, where sample size allows reasonable and reliable interpretations of the analyses.

Data selection Data selection involved two stages. At Stage 1, compositions were selected according to three requirements concomitantly. First, they should represent five writing topics - two ‘personal’ and three ‘impersonal’, with a good range of scores. Secondly, to ensure high inter-rater reliability, a composition should receive the same score3 from both raters. Altogether 201 compositions were selected. One was rejected because of its numerous spelling mistakes. At Stage 2, all tape-recorded interviews of the 201 candidates who also took the speaking test were grouped by their composition scores (e.g. 73), then within each group a range of interviews with different oral scores were selected. Altogether we selected 26 interviews. One interview proved too difficult to transcribe and was discarded.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

RESEARCH PROCEDURES

G. YU

243

Table 1: Gender, composition topic, test purpose and first language crosstabulation Composition topic IDa L1

Test purpose

1

Tagalog/Filipino College Professional

Female Female 14 Male 1 Other purpose Female

Chinese

College

Graduate

Female Male Female Male Female

College Professional Other purpose Graduate

Male Female Female Female

Professional

Russian

Farsi/Persian

College Professional

Female Female Male Other purpose Female Junior college Male

2

3

4

5

Sub-total

2

12 2

7 1 2

1 12 1 1

1 47 5 4

1

4 6

2 3 3

7 9 6 3 1

1 3 1 1

2 7 1 2

1 1 1 1

3 5 1 2 1

1

2

1 1 1

2

2

2

1

1

2

2 1

1

1 1

(continued)

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

As a result, the usable data had 200 compositions and 25 interviews. 155 compositions were written by female candidates (77.5%) and 45 by male (22.5%). Half of them had impersonal topics (female = 86, male = 14) and the other half (female = 69, male = 31) personal topics. The two main purposes for taking MELAB written test were for ‘college’ admission (n = 48) and ‘professional’ certification (n = 129). Most interviewees (21/25) did the test for professional certification (mainly nursing), with only three of them for college admission and one for other purpose. Most interviewees (22/25) were females. In terms of candidates’ L1 in the composition data, Tagalog/ Filipino, Chinese, Farsi/Persian, Russian, were the four main groups. In the interview data, 1/3 of the candidates’ L1 was Tagalog/Filipino, followed by Chinese. The means of compositions and speaking scores were 77.1 and 7, respectively. The selected sample was reasonably representative of MELAB candidates during this period in terms of test purpose, gender, L1 and test performance (see MELAB Technical Manual 2003 and Johnson 2005) (Table 1).

244 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

Table 1: Continued Composition topic IDa L1

Test purpose

Other L1b

College Professional Other purpose

Junior college a

Female Male Female Male Female Male Female Male Male

1 11

2

3

4

5

Sub-total

3 1 11 4 3

2 2 10

4 3 4 2 1 2

6 4 11 2

16 10 47 8 4 2 1 2 1

1 1

1 1

Topics 1–3: impersonal; Topics 4 and 5: personal. ‘Other L1’ includes 34 different languages.

b

Preparation of transcripts and running vocd The researcher processed the handwritten compositions in MS Word, without trying to correct all grammatical or spelling mistakes. However, a spelling mistake would be changed according to the following principle: If a mis-spelt word was correct elsewhere, then it was corrected. Otherwise, the mis-spelt word was entered as it was. However, if the same word was spelt incorrectly differently in a few places (2), one of the wrong spellings was randomly chosen to replace all the others, because entering two or more differently mis-spelt words of the same word could potentially increase the value of D. When transcribing the interviews, utterances of the examiner and non-words such as laughter and pause markers were omitted, since no pragmatic turntaking analysis was to be performed. Apart from that, the interviewees’ utterances were transcribed word by word by a professional and double-checked by the researcher. Some researchers treated all inflections of the same word as a single type when measuring learners’ lexical knowledge (e.g. Laufer 1991; McClure 1991; Engber 1995), some distinguished ‘inflected forms’, ‘stem forms’ and ‘root forms’ to understand learners’ morphological development (e.g. Dura´n et al. 2004), based on the assumption that treating different inflections of the same word as separate types might confound lexical with grammatical knowledge. These approaches may fit well with the particular research foci of these studies. This present research, however, considered the inflections of the same word as different types, for three reasons: (i) the main purpose of this research is to examine lexical diversity as a quality indicator of an end-product,

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Graduate school

1

G. YU

245

RESULTS AND DISCUSSION An overview of the data The compositions had 61790 tokens and 4325 types in total. The average word and sentence lengths were 4.3 letters (std. deviation = 0.36) and 21.6 words (std. deviation = 7.38), respectively. The length of the compositions varied considerably5 from 123 to 735 words (mean = 309, std. deviation = 96.9). D ranged from 32 to 161 (mean = 73, std. deviation = 18.9). Dura´n et al. (2004) noted in their sample that D of adult academic writings ranged from about 70 to 120 (mean = 91). Judging from Dura´n et al.’s statistics, it is quite reasonable to speculate that on average MELAB candidates may not have reached the academic writing level, except for a very few of them. The interviews lasted from 204 to 1183 s6 (mean = 510, std. deviation = 201), and had 210 to 2163 words (mean = 678, std. deviation = 410). Altogether the interview data had 16961 tokens and 1981 types, and most of the words (61%) consisted of two to four letters. It had far more short words (mean = 618, std. deviation = 358) than long words (mean = 60, std. deviation = 56). D of the interviews ranged from 51 to 106 (mean = 74, std. deviation = 15).

Addressing the five research questions Research Question 1: Whether and to what extent the lexical diversity of the written discourses and the scores assigned to them are related? Simple linear regression analysis was conducted with D as the predictor7 and the overall quality score as the dependent. The Pearson correlation (r = 0.294, p < 0.001, N = 200) indicated that compositions with a higher D tended to have received a higher score. The case-wise diagnostics identified two outliers (No. 95 and No. 127) outside three standard deviations (see further discussion later). After removing the two outliers, the Pearson correlation increased to

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

rather than the lexical knowledge of the producer as such, (ii) treating different inflections of the same word as a single type would not have had any appreciable effect on the results since there were only few inflections in the data according to the initial word frequency analyses (available from author upon request), and (iii) the MELAB rating scale particularly highlights that candidates need to demonstrate ‘accurate morphological (word forms) control’ as one key quality indicator of test performance. Every text had more than 50 words, thus meeting the minimum sample size requirement to compute a valid D. Each text was subject to 15 times of vocd analyses; a slightly different D was reported each time,4 therefore, the average of them was used as its final D. In addition, freq and wdlen were run to make a word list and its frequency statistics, and word length (i.e. number of characters), respectively. WordSmith Tools (Scott 2004) were also used to measure sentence length (i.e. number of words in a sentence) of the compositions.

246 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

0.33 (t = 4.895, p < 0.001). In other words, D was able to account for around 11% of the variances in the overall quality ratings. This significant correlation is high, considering the myriad of other factors that may affect raters’ judgement, e.g. handwriting quality and syntactical complexity. It is desirable to include other lexical features such as average word and sentence lengths, number of long and short words, and the length of a composition in one single regression analysis in order to understand the differential predictability of D and other aforementioned lexical features. However, serious collinearity due to the strong correlations between the independent variables was noted, in particular between D and number of types (r = 0.522, p < 0.001), as expected, because D is related mathematically to the number of types. Therefore, a series of regression analyses were run, instead, using D and one of the aforementioned lexical features as the two predictors and the overall quality rating as the dependent (Table 2). No serious collinearity was noted. As shown in Table 2, D was significant in all models except Model 2 where the number of types seemed to be much more important to predict the overall quality score. In Model 1, the number of tokens and D seemed to be equally important. Compared with ‘word length’, ‘sentence length’ and ‘number of short words’ in Models 3, 4 and 6 respectively, D was more able to predict the overall quality score. However, compared with ‘number of long words’, D was a less powerful predictor (Model 5). Although the number of long words is apparently linked with the number of tokens of a composition, it seems that what words as well as how many words— though in a less degree—were used by the writers may have affected raters’ judgement. The outliers observed in the various regression analyses (Table 2) are particularly noteworthy. The D values of compositions No. 95 and No. 127 were 102.4 and 68.6, respectively; while these two compositions received 60 and 97 for their overall quality respectively. Further examination of the compositions per se showed that No. 127 was beautifully written by an English nativespeaker taking the test for professional certification. However, No. 95 not only deviated from the topic but also had many grammatical mistakes, although it did seem that several nonsensical sentences in the composition could have increased its D. The two outliers demonstrated that D alone may not capture the overall quality of the compositions because it can only identify how often different words are used but not how they are used or where they are used, or what the different words are. This could have significant implications for the development of automated writing evaluation systems employing lexical diversity as an important parameter. It seems that bad writers may well be able to trick automated evaluation systems, or at least the parameter of lexical diversity of the systems, to get a good score. This finding seems to confirm but also contradict at the same time what Shermis and Burstein (2003) argued: a bad essay can get a good score in the automated evaluation systems, but it takes a good writer to produce the bad essay to get a good score and it is almost impossible for a bad writer to produce a bad essay that gets the

5, R2 = 0.217 Outliers: No. 25 95, 110, 126, 127 6, R2 = 0.135 Outliers: No. 126, 127

4, R2 = 0.144

3, R2 = 0.118 Outliers: No. 95, 127

2, R2 = 0.219 Outliers: as above

64.81 0.09 0.02 63.81 0.03 0.08 60.98 0.11 1.87 64.11 0.11 0.23 65.45 0.09 0.15 65.54 0.10 0.02

1, R2 = 0.156 Outliers: No. 110, 126, 127

Constant D Tokens Constant D Types Constant D Average word length Constant D Average sentence length Constant D Number of long words Constant D Number of short words

B

2.06 0.02 0.01 1.91 0.03 0.01 5.59 0.03 1.33 2.30 0.03 0.06 1.74 0.02 0.03 2.10 0.02 0.01

Std. error

0.27 0.21

0.26 0.33

0.30 0.24

0.31 0.10

0.08 0.42

0.26 0.26

Beta

Standardized coefficients Sig.

.000 .000 .000 .000 n.s. .000 .000 .000 n.s. .000 .000 .000 .000 .000 .000 .000 .000 .002

t

31.51 3.80 3.84 33.39 1.12 5.62 10.90 4.50 1.40 27.85 4.51 3.63 37.66 4.03 5.03 31.21 3.96 3.16

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Model

Unstandardized coefficients

Table 2: Coefficients of D and another lexical feature with overall quality rating of compositions

60.75 0.04 0.01 60.04 0.02 0.05 49.94 0.06 0.75 59.57 0.06 0.10 62.02 0.05 0.09 61.39 0.05 0.01

Lower bound

68.86 0.14 0.07 67.58 0.08 0.10 72.01 0.16 4.48 68.64 0.16 0.35 68.87 0.14 0.21 69.68 0.14 0.03

Upper bound

95% Confidence interval for B

G. YU 247

248 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

Gender Although the lexical diversity of compositions by females and males did not differ significantly (female = 73.2, n = 155; male = 70.4, n = 45), the regression analyses on data of females and males separately demonstrated that the correlation between D and overall quality rating was much more pronounced in the compositions by males (R2 = 0.34) than by females (R2 = 0.05). D alone could explain 34% of the variances in the overall quality ratings of compositions by males, but only 5% for females. Applying the six regression models (see Table 2) on the male and female data separately, similar patterns were also noticed. The R2 ranged from 35.5% to 47.5% for compositions by males. However, the explaining power of D was much less for compositions by females (R2 from 5.2% to 16.5% only).

L1 background Due to the small number of candidates from other L1 background, only the two largest L1 groups—Filipinos and Chinese—were compared. Compositions by the two groups did not have significant differences in D (meanFilipinos = 70.6, n = 57; meanChinese = 69.5, n = 26) or overall quality ratings (meanFilipinos = 77.7; meanChinese = 75.9). The simple linear regression analyses on the Filipino data demonstrated that D was not able to make any significant prediction of the overall quality ratings. Even within the six models together with other lexical features, no significant correlation was found between the lexical features and the overall quality ratings. For the Chinese sample, D had no significant correlation with overall quality rating, either. However, the regression analyses on the Chinese sample using the six models (see Table 2) presented a very different picture from the Filipino data. D together with the other lexical features, in particular, numbers of types, tokens, short and long words, and average sentence length, were able to predict from 33.5%

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

good score. Except for No. 95, all the other outliers in the regression analyses received very high overall ratings: No. 126 and No. 127 received the maximum possible score 97, No. 110 was assigned 95, and No. 25, 93 points. But their D values were just around the sample mean. No. 126 and No. 127, both written by English native speakers, probably raise the question of the applicability of D for different groups of participants, particularly when they were treated as if from the same population, in this case, both as EFL learners. It also raises the question of whether there were differences in D of the compositions by native speakers and EFL learners, between males and females, between L1 groups, between test purposes, and between different topics (see RQ2), and more importantly, whether the findings on the whole sample hold true for each sub-group within the sample. Therefore, the six models of the regression analyses as for the whole sample (see Table 2) were also applied on data of each sub-group.

G. YU

249

up to 43.6% of the variances of the overall quality ratings. However, in none of the models D was significant. It seems that the length of the compositions written by the Chinese candidates were much more important than D in predicting the overall quality scores. Further regression analyses on the rest of the sample (n = 117) confirmed that the significant relationship between D and overall quality rating of the whole sample may well be largely due to the compositions by candidates from other L1 backgrounds (R2 = 0.11).

Purpose of taking the test Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

There was significant difference (t = –1.976, df = 175, p < 0.05) in D of compositions written by candidates taking the test for professional certification (meanprof = 73.62, n = 129) and those for college admission (meancollege = 67.47, n = 48). A series of similar regression analyses was also run as for gender and L1 background. It was found that the R2 value for the college admission group was about twice of that for professional certification group, except for when D was the only predictor where the R2 for professional certification group was similar to that for the college admission group. Similar analyses were applied to understand the effects of topic types (personal vs. impersonal). In most of the six models, the R2 values for impersonal compositions were about half of those for personal compositions (see RQ2 for further analyses). In summary, these analyses evidence that D was a significant predictor of the overall quality rating of compositions. However, other lexical features such as the number of types, tokens, short and long words, and average word and sentence length may also exert similar effects. In particular, the number of types (see Model 2, Table 2) and the number of long words (see Model 5, Table 2) seemed to be the other two most illuminative indicators, besides D, for the overall quality of the compositions. Together with D, they were able to predict a fairly large amount of the variances in overall quality rating. It is also interesting to note that the significant correlations between D and overall quality rating did not always hold true for sub-groups of the sample. For example, D was a much less powerful variable to predict the overall quality ratings of compositions written by females than males. The non-significant relationship between D and overall quality ratings of compositions by Filipinos or Chinese demonstrates that the significant correlation which was found in the whole sample was largely attributable to significant correlations in compositions written by other L1 candidates. The correlation of D and overall quality was also stronger for compositions written by candidates for college admission than for professional certification purposes, though D values of the compositions for professional certification were significantly higher than those for college admission. Similarly, the correlations between D and overall quality of personal compositions were stronger than those for impersonal compositions. However, due to the small sample size and the significant correlations between D and the other lexical features, it is difficult to tease out

250 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

the combined effects of the above factors simultaneously on the overall quality ratings. These findings have a clear message for test validation and development of automated evaluation systems using lexical diversity parameters—one size does not fit for all. An approach highly appropriate for one sub-group of the sample may be not relevant at all to the other group. The use of D to predict the overall quality of a discourse therefore should take into account simultaneously various contextual factors, e.g. the characteristics of candidates and tasks. Research Question 2: Whether and to what extent the lexical diversity of the written discourses is affected by the writing topics? Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The MELAB sample consisted of three impersonal and two personal topics. The independent samples t-tests showed that there was a statistically significant difference in D (t = 2.41, df = 198, p < 0.05) between personal (mean = 69.4, std. deviation = 16.4) and impersonal compositions (mean = 75.7, std. deviation = 20.6). Furthermore, ANOVA statistics present a more complex picture. Overall there existed significant differences in D between the five actual topics (F4, 195 = 9.064, p <.001). The Scheffe´ test indicated that compositions of Topic 2 (impersonal) and Topic 5 (personal) formed a homogenous sub-set, and Topics 1, 3, 4, 5 formed another homogenous subset ( =.05). The t-test and ANOVA statistics above demonstrate significant difference in D existed between different writing topics and between personal and impersonal topic types. Even after controlling the effects of candidates’ writing abilities (i.e. composition score), and overall language proficiency (i.e. Final MELAB score), the effects of writing topics and topic types still held statistically significant, as evidenced in four full-factorial univariate ANOVAs on D as the dependent, and writing topics or topic types as the between-subject factor, and composition or Final MELAB score as a covariate. The key statistics of the models are as follows: Model A (Fcomp = 21.83, p <.001; Ftopics = 9.96, df = 4, p <.001; R2 = 0.242), Model B (Ffinal = 20.93, p <.001; Ftopics = 8.48, df = 4, p <.001; R2 = 0.239), Model C (Fcomp = 19.7, p <.001; Ftopictype = 6.78, p <.001; R2 = 0.117), and Model D (Ffinal = 25.7, p <.001; Ftopictype = 7.97, p <.001; R2 = 0.141). Close investigation of the actual topics and the characteristics of the candidates help to shed some light on this. Because the vast majority of the sample took the test for professional certification, mainly nursing, Topic 2 (i.e. plastic surgery) could well be a familiar and easier topic to most of them, which might have increased the chance for them to produce compositions of higher lexical diversity because of topic familiarity, knowing more words in relation to the topic, being more confident and having more planning time as a result (see O’Loughlin 1995). This finding has significant implications for the design of writing topics. Although a familiar topic may help to elicit an adequate large sample, issues of test bias, fairness and ethicality are

G. YU

251

equally important to consider, if the candidates are not from a homogenous group. Research Question 3: Whether and to what extent the lexical diversity of the spoken discourses and the scores assigned to them are related?

Research Question 4: Whether and to what extent the lexical diversity of the written and the spoken discourses are related to the candidates’ overall language proficiency? The Final MELAB score, calculated as the average of the scores on composition, listening and GCVR (Grammar, Cloze, Vocabulary and Reading) subtests, was used as a proxy of candidates overall language proficiency. As reported in RQ2 on the effects of topics and topic types on lexical diversity, Final MELAB score was found to have significant effect on lexical diversity of compositions. Simple regression analyses using Final MELAB as the single predictor and D as the dependent demonstrated that Final MELAB was able to account for 9.3% of the variances in D of compositions (t = 4.497, p < .001, N = 199) and 24.7% of the variances in D of interviews (t = 2.748, p < .01, n = 25), respectively. Using scores on Listening and GCVR (i.e. excluding scores on compositions) yielded similar findings—9% of the variances in D of compositions were explained by Listening and GCVR scores (t = 4.413, p < .001, N = 200), and 20.9% of the variances in D of interviews were explained (t = 2.413, p < .05). These findings were in the same pattern as those of RQ1

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The much smaller sample size (n = 25) of the interviews limits the methods of analyses. Tabachnick and Fidell (2003: 117), quoting Green (1991), suggest a formula for calculating sample size requirements (n  50+8m for testing multiple correlation, and n  104+m for testing individual predictors, where m is the number of predictors). Ideally, more interviews are required for reliable regression analyses using two predictors. To make the best use of the data, simple regression analysis having D as the single predictor and test score as the dependent variable was used, instead. It was found that lexical diversity and the overall score of interviews were significantly related (r = 0.484, p <.01). In other words, around 23.4% of the variances in the oral score could be explained by D of the interviews. Considering many other potential factors affecting interviewers’ judgement of test performance, e.g. fluency, pronunciation and turn-taking strategies, this 23% is substantial. Further regression analyses using another lexical feature a time showed that the number of tokens (R2 = 0.167, t = 2.146, p < .05) and the number of types (R2 = 0.22, t = 2.55, p < .05) were also significant predictors, so was the number of short words (R2 = 0.167, t = 2.149, p < .05) unexpectedly.8 Unlike the composition data, the number of long words was at a borderline of significance (R2 = 0.14, t = 1.937, p < .0655) for interviews. Overall, D could explain larger variances of the overall quality scores of interviews than other lexical features. In addition, the magnitude of the predictability of D for interviews also seems to be bigger than that for written compositions.

252 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

(relationship between D and scores on compositions) and RQ3 (relationship between D and scores on interviews), largely because the candidates’ writing, speaking, listening and GCVR and Final MELAB scores were highly correlated as would be expected (Pearson correlations ranged between 0.73 and 0.99). In all cases D had substantial and significant correlations with the producers’ writing and speaking abilities, and overall language proficiency, and this is particularly prominent in the interview data. However, it should be noted that independent measures of language abilities would be more desirable for a better understanding of the relationships between lexical diversity and language proficiencies.

In the sample, there were 25 candidates whose compositions and interviews were available. All the five writing topics were covered by the interviewees. Seventeen of them wrote impersonal topics (seven, four and six interviewees for Topics 1, 2 and 3, respectively) and eight personal topics (two and six interviewees for Topics 4 and 5, respectively). The lexical diversity of the 25 compositions and interviews was significantly correlated (r = 0.448, p < .05), and the mean difference in Ds was very small (mean Dcomp = 76.29, std. deviation = 27.81; mean Dinterviews = 74.11, std. deviation = 14.73; t = 0.437, n.s.). This finding suggests that the lexical diversity of candidates’ speaking and writing outputs was not only significantly positively related but also approximately at the same level. The implications of this stability of lexical diversity of candidates’ performances in written compositions and face-to-face interviews are twofold. First, it indicates that candidates may write and speak in a similar manner in terms of the degree of lexical diversity of their outputs, even in different contexts of time pressure. The writing task does not deprive candidates from the chance of planning and (re)organizing words and sentences. In such a way, the lexical diversity of the writings might be increased. The spontaneous face-to-face interviews do not have the luxury of pre-task planning, which may decrease the chance for achieving high lexical diversity. However, spontaneous interviews may have more ‘scrambled’ utterances (as composition No. 95) than the written compositions, which may well have increased D of the interviews and counter-balanced the writing tasks’ advantages affording the writers for achieving higher lexical diversity. Quite contrary to O’Loughlin (1995), this non-significant difference in lexical diversity between writing and speaking suggests that lexical diversity may not be affected by task types (written compositions or spontaneous interviews) or other task characteristics such as pre-task planning and time pressure. However, more empirical studies are needed to support the above speculation. Secondly, this kind of stability provides evidence for the validity and importance of including lexical diversity as a salient quality indicator in the MELAB rating scales on the one hand, and the applicability and usefulness of D to

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Research Question 5: Whether and to what extent the lexical diversity of the written and the spoken discourses are related?

G. YU

253

measure lexical diversity of both writing and speaking task performances on the other. It is quite clear from the main findings of RQ1, RQ3, and RQ4 that D and the overall quality ratings were statistically significantly related for both writing and speaking performances. It is nevertheless still a mystery why D which was at approximately the same level for both writing and speaking performances was more able to predict speaking than writing performances. More research using larger interview samples is needed to explore these dynamics further.

CONCLUSION Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

This article has examined (a) the relationships between lexical diversity and overall quality of MELAB candidates’ writing and speaking task performances and their general English language proficiency, (b) the relationships between candidates’ writing and speaking lexical diversity, and (c) to what extent different topics and topic types (impersonal vs. personal) of the writing prompts affect lexical diversity of the compositions. The analyses of the sample from MELAB 2004–2005 data evidenced that lexical diversity D had statistically significant and positive correlations with candidates’ general English language proficiency and the judgement of the overall quality of writing and speaking task performances. It was also noted that lexical diversity was more able to predict candidates’ speaking than writing task performances although lexical diversity of speaking and writing performances was approximately at the same level. However, the magnitude of the correlations between D and overall quality of written compositions varied to a great extent across the different subgroups of the sample, in terms of gender, L1 background, test taking purpose, and the writing prompts. For instance, the correlation for the males was nearly twice that for the females, although the two subgroups did not differ significantly in their lexical diversity, a finding quite contrary to Singh’s (2001) on gender differences in lexical diversity measures based on word-frequencies. For the two largest L1 groups (Filipino and Chinese) D was not a significant predictor of the overall quality ratings of their written compositions, while for the rest of the sample such correlation was statistically significant. One possible explanation may be that there exist some language distance effects (Elder and Davies 1998) on the habitual writing practice of different L1 background and consequentially on lexical diversity of compositions. However, research with larger samples is needed to understand why there were no significant correlations between D and overall quality ratings of the Filipino and Chinese compositions. In addition, the two English native-speaker outliers (No. 126 and 127) noticed in the regression analyses were most telling. Close examination of the compositions per se would not suggest they were poorly written at all, however, their Ds were lower than predicted. It might be purely by chance that these two outliers happened to be English native speakers’ (there were altogether eight compositions written by English native speakers

254 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

in the sample), but it does suggest that it may be necessary to apply different lexical diversity measures on compositions by native and non-native speakers in order to understand the differences between these two groups. The significant differences in lexical diversity found between native speakers and non-native speakers in other studies (e.g. Linnarud 1986; Jarvis 2002) might be due to the different conceptualizations and measurements of lexical diversity from this current research. These complexities seem to suggest that the one-size-fits-all approach in (a) validating writing tasks, (b) developing automated evaluation systems using lexical diversity measures and (c) applying such systems to compositions written by candidates of different personal and linguistic characteristics and background are not fully justified. Different topics and topic types of the writing prompts exerted significant effects on the lexical diversity of the compositions, even after controlling the effects of the candidates’ writing abilities and overall language proficiencies. Compositions on impersonal topics had significantly higher lexical diversity than personal topics. Higher lexical diversity was achieved when candidates were highly familiar with the topic. The significant effect of topics and topic types on lexical diversity echo the suggestion made by Vermeer (2000: 79) ‘control of tasks over informants is a prerequisite for comparing different texts, so that the kinds of topics are similar for all informants’, concurred by Owen and Leonard (2002: 936). In addition, in MELAB context it would be interesting to look at how candidates make their decision in choosing a writing topic9 (e.g. Jennings et al. 1999) and to what extent the topic selection affects lexical diversity of compositions. For example, had they chosen a different writing topic, would they have produced a composition of similar lexical diversity as it now stands? It should be fully acknowledged that although lexical diversity is an essential quality indicator of written compositions and spontaneous interviews as evidenced in this research, it is not the only quality indicator of task performance, various other facets (e.g. handwriting and quality of presentation) may affect raters’ judgement. Ceteris paribus, raters tend to favor well handwritten compositions which in turn receive higher scores than their poorly handwritten counterparts. Although the raters agreed well for the sample of this research (see Note 3), it is nevertheless important to bear in mind that it is far from certain that it was because of the same list of qualities of a composition that made the raters give the same/similar scores, since raters’ decision-making is ‘a somewhat indeterminate process’ (p. 246) and may involve a series of reconciliations of their intuitive impression and interpretation of the compositions, the specific features of the compositions, and the wording of the official rating scales (Lumley 2002). Furthermore, raters may be influenced by not only the order a composition was marked, i.e. rating context variables (e.g. Daly and Dickerson-Markman 1982), but also handwriting and quality of presentation (e.g. Briggs 1970; Chase 1979; Hughes et al. 1983; Powers et al. 1994). Further research studies are apparently needed

G. YU

255

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

to control the effects of the quality of handwriting and the order of the compositions to be marked in order to have a much better understanding of the relationships between lexical diversity and overall quality judgement of the compositions. However, even after controlling for the extraneous factors such as the effects of writing prompts and the handwriting and quality of presentations, lexical diversity measures, like D, are not able to capture what lexical items are actually used and how. As Vermeer (2000) suggested: ‘Instead of counting how many types and tokens pop up in the data set, a more valid measure of lexical richness might be to relate the words in the data to their difficulty, measured by their frequency (or frequency classes) in corpora’ (p. 79), an approach similar to the Lexical Frequency Profile (LFP) proposed by Laufer and Nation (1995). It seems that an important aspect for future studies, in addition to using lexical diversity measures like D, is to compare and contrast what actual words are used, the frequency of basic and advanced lexical items (e.g. Daller et al. 2003) and the rate of open and closed lexical items in a discourse by different groups of candidates, e.g. in terms of gender, L1, writing topics, writing abilities, and general language proficiencies. However, as Meara (2005; Meara and Bell 2001) showed that LFP is also problematic (but see Laufer 2005 for her different views), because it may not be sensitive enough to ‘pick up modest changes in vocabulary size’ although it ‘works best when the groups being compared have very disparate vocabulary sizes’ (Meara 2005: 32). A single, perfect measure of lexical diversity fit for all research purposes may be just like the Holy Grail (Malvern et al. 2004), but it is essential to unambiguously define what the lexical diversity indicators are and how they are measured in each individual research study. For example, whether raw or lemmatized data are used may make a clear difference in measuring lexical diversity of a discourse (Richards and Malvern 2007). As noted in the literature review, the different nomenclatures, conceptualizations and measurements of lexical diversity across a variety of fields in applied language studies make it difficult to synthesize research findings and apply them into practice. It is also demonstrated in the current article that the use of different measures (e.g. TTR and D) of the seemingly single concept of lexical diversity may well produce different and conflicting research findings. However, this should not be interpreted exclusively as one indicator is better than the other; on the contrary it urges the importance and necessity of understanding lexical diversity from multiple perspectives and measures (see also Nation 2007; Richards and Malvern 2007). Furthermore, as noted by Daller and Phelan (2007) human raters may attend to different aspects of lexical diversity when judging the overall quality of a discourse (see Lumley 2005 for further complexities of rater decision-making processes), it is therefore essential to compare like with like when extrapolating human rater judgement with computerized lexical diversity measures.

256 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

ACKNOWLEDGEMENTS I would like to thank University of Michigan for supporting me to undertake this research through its Spaan Fellowship for Studies in Second or Foreign Language Assessment, especially Dr Jeffrey Johnson and his colleagues in the Testing Division of the English Language Institute for assisting data selection. My thanks are also due to Professor Pauline Rea-Dickins and Dr Richard Kiely of University of Bristol, Professor Jeanine Treffers-Daller and Dr Helmut Daller of University of the West of England, and the three anonymous reviewers for their insightful comments on earlier drafts of this article. However, any error remains mine. An earlier version of the article was presented at the 40th annual conference of British Association for Applied Linguistics (Edinburgh, 6–8 September 2007). Conflict of interest statement: None declared.

1

2

3

4 5

6

In some studies, this is then defined as the producer’s lexical or vocabulary knowledge or size. D may not be completely free of sample size, as shown in some studies (e.g. Owen and Leonard 2002; McCarthy and Jarvis 2007), but it is not a function of sample size as TTR and other measures are (Malvern et al. 2004: 73–74). In fact, 6 of the selected compositions did not receive identical scores; they were just the average of the two adjacent ratings. This is because the computation of D is based on random sampling of a text. This wide range may be partly due to the task direction which did not specify the number of words required. According to MELAB Information Bulletin, candidates are expected to ‘have a 10–15 minute conversation with the local examiner’ (p.10). The data do not seem to conform to the specifications, but reassuringly there is no significant correlation between ratings and duration of the interviews.

7

8

9

Exactly in the same manner, the predictability of TTR for the overall quality ratings was tested. TTR was around half of the predictability of D. Further regression analyses using both TTR and D as the predictors confirmed that D was more able than TTR (Beta = 0.317, t = 3.875, p < 0.001 for D; Beta = 0.023, t = 0.278, n.s. for TTR). Therefore, all the subsequent analyses used D as the main lexical diversity measure, and TTR per se was not used. Nevertheless, D is mathematically related to number of types and tokens, therefore number of types and tokens as components of TTR were included instead in the subsequent analyses. My thanks are due to one of the reviewers for pointing out this. It is probably due to the fact that the means of ‘tokens’ and ‘short words’ were very close, as well as their standard deviations. MELAB writing task provides two topics/prompts for candidates to choose.

REFERENCES Arnaud, P. J. L. 1984. ‘The lexical richness of L2 written productions and the validity of

vocabulary tests’ in T. Culhane, C. KleinBraley, and D. K. Stevenson (eds): Practices

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

NOTES

G. YU

corpora’ in J. M. Kirk (ed.): Corpora Galore: Analyses and Techniques in Describing English. Rodopi, pp. 249–57. Green, S. B. 1991. ‘How many subjects does it take to do a regression analysis?,’ Multivariate Behavioral Research 26: 499–510 Guiraud, P. 1954. Les caracte`res statistiques du vocabulaire. Presses universitaires de France. Guiraud, P. 1960. Proble`mes et me´thodes de la statistique linguistique. D. Reidel. Ha¨rnqvist, K., U. Christianson, D. Ridings, and J. G. Tingsell. 2003. ‘Vocabulary in interviews as related to respondent characteristics,’ Computers and the Humanities 37: 179–204. Herdan, G. 1960. Type-Token Mathematics: A Textbook of Mathematical Linguistics. Mouton. Hoover, D. L. 2003. ‘Another perspective on vocabulary richness,’ Computers and the Humanities 37: 151–78. Ho¨weler, M. 1972. ‘Diversity of word usage as a stress indicator in an interview situation,’ Journal of Psycholinguistic Research 1: 243–48. Hughes, D. E., B. Keeling, and B. F. Tuck. 1983. ‘Affects of achievement and handwriting quality on scoring essays,’ Journal of Educational Measurement 20: 65–70. Hyltenstam, K. 1988. ‘Lexical characteristics of near-native second-language learners of Swedish,’ Journal of Multilingual and Multicultural Development 9: 67–84. IELTS. 2007. International English Language Testing System Handbook 2007. Jarvis, S. 2002. ‘Short texts, best-fitting curves and new measures of lexical diversity,’ Language Testing 19: 57–84. Jennings, M., J. Fox, B. Graves, and E. Shohamy. 1999. ‘The test-takers’ choice: An investigation of the effect of topic on language test performance,’ Language Testing 16: 426–56. Johnson, J. S. 2005. MELAB 2004: Descriptive Statistics and Reliability Estimates. University of Michigan: English Language Institute. Johnson, R. L. 1979. ‘Measures of vocabulary diversity’ in D. E. Ager, F. E. Knowles, and J. Smith (eds): Advances in Computer-Aided Literary and Linguistic Research. University of Aston: Department of Modern Languages, pp. 213–27. Klee, T., S. F. Stokes, A. M. Y. Wong, P. Fletcher, and W. J. Gavin. 2004. ‘Utterance length and lexical diversity in Cantonese-speaking children with and without

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

and Problems in Language Testing. University of Essex, pp. 14–28. Bradac, J. J., J. Bowers, and J. A. Courtright. 1979. ‘Three language variables in communication research: Intensity, immediacy, and diversity,’ Human Communication Research 5: 257–69. Briggs, D. 1970. ‘The influence of handwriting on assessment,’ Educational Research 13: 50–55. Carroll, J. B. 1964. Language and Thought. Prentice-Hall. Chase, C. I. 1979. ‘The impact of achievement expectations and handwriting quality on scoring essay tests,’ Journal of Educational Measurement 16: 39–42. Chodorow, M. and J. Burstein. 2004. Beyond Essay Length: Evaluating e-rater’s Performance on TOEFL Essays (Research Report No. 73). Educational Testing Service. Ciani, A. J. 1976. ‘Syntactic maturity and vocabulary diversity in the oral language of first, second, and third grade students,’ Research in the Teaching of English 10: 150–56. Crystal, D. 1982. Profiling Linguistic Disability. Edward Arnold. Daller, H. and D. Phelan. 2007. ‘What is in a teacher’s mind? Teacher ratings of EFL essays and different aspects of lexical richness’ in H. Daller, J. Milton, and J. Treffers Daller (eds): Modelling and Assessing Vocabulary Knowledge. Cambridge University Press, pp. 234–44. Daller, H., R. Van Hout, and J. TreffersDaller. 2003. ‘Lexical richness in the spontaneous speech of bilinguals,’ Applied Linguistics 24: 197–222. Daly, J. A. and F. Dickerson-Markman. 1982. ‘Contrasts effects in evaluating essays,’ Journal of Educational Measurement 19: 309–16. Dura´n, P., D. Malvern, B. Richards, and N. Chipere. 2004. ‘Developmental trends in lexical diversity,’ Applied Linguistics 25: 220–42. Educational Testing Service. 2005. TOEFL iBT Scores. Elder, C. and A. Davies. 1998. ‘Performance on ESL examinations: Is there a language distance effect?,’ Language and Education 12: 1–17. Engber, C. 1995. ‘The relationship of lexical proficiency to the quality of ESL compositions,’ Journal of Second Language Writing 4: 139–55. Granger, S. and M. Wynne. 2000. ‘Optimising measures of lexical variation in EFL learner

257

258 LEXICAL DIVERSITY IN WRITING AND SPEAKING TASK PERFORMANCES

O’Loughlin, K. 1995. ‘Lexical density in candidate output on direct and semi-direct versions of an oral proficiency test,’ Language Testing 12: 217–37. Owen, A. J. and L. B. Leonard. 2002. ‘Lexical diversity in the spontaneous speech of children: With specific language impairment: Application of d,’ Journal of Speech Language and Hearing Research 45: 927–37. Panas, E. and A. N. Yannacopoulos. 2004. ‘Stochastic models for the lexical richness of a text: Qualitative results,’ Journal of Quantitative Linguistics 11: 251–73. Powers, D. E., M. E. Fowles, M. Farnum, and P. Ramsey. 1994. ‘Will they think of less of my handwritten essay if others words process theirs? Effects on essay scores of intermingling handwritten and word-processed essays,’ Journal of Educational Measurement 31: 220–33. Read, J. 2000. Assessing Vocabulary. Cambridge University Press. Richards, B. 1987. ‘Type/token ratios: What do they really tell us?,’ Journal of Child Language 14: 201–09. Richards, B. and D. Malvern. 2007. ‘Validity and threats to the validity of vocabulary measurement’ in H. Daller, J. Milton, and J. Treffers Daller (eds): Modelling and Assessing Vocabulary Knowledge. Cambridge University Press, pp. 79–92. Sankoff, D. and R. Lessard. 1975. ‘Vocabulary richness: A sociolinguistic analysis,’ Science 190: 689–90. Scott, M. 2004. Wordsmith Tools. Oxford University Press. Shermis, M. D. and J. Burstein. 2003. ‘Introduction’ in M. D. Shermis and J. Burstein (eds): Automatic Essay Scoring: A Cross-Disciplinary Perspective. Lawrence Erlbaum. Sichel, H. S. 1986. ‘Word frequency distribution and type-token characteristics,’ Mathematical Scientist 11: 45–72. Silverman, S. and N. B. Ratner. 2002. ‘Measuring lexical diversity in children who stutter: Application of vocd,’ Journal of Fluency Disorders 27: 289–304. Singh, S. 2001. ‘A pilot study on gender differences in conversational speech on lexical richness measures,’ Literary and Linguistic Computing 16: 251–64. Sokolova, M., M. Shah, and S. Szpakowicz. 2006. ‘Comparative analysis of text data in successful face-to-face and electronic

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

specific language impairment,’ Journal of Speech Language and Hearing Research 47: 1396–410. Laufer, B. 1991. ‘The development of L2 lexis in the expression of the advanced learner,’ Modern Language Journal 75: 440–48. Laufer, B. 2005. ‘Lexical frequency profiles: From Monte Carlo to the real world,’ Applied Linguistics 26: 582–88. Laufer, B. and P. Nation. 1995. ‘Vocabulary size and use – lexical richness in L2 written production,’ Applied Linguistics 16: 307–22. Linnarud, M. 1986. Lexis in Composition: A Performance Analysis of Swedish Learners’ Written English. Liber Fo¨rlag Malmo¨. Lumley, T. 2002. ‘Assessment criteria in a largescale writing test: What do they really mean to the raters?,’ Language Testing 19: 246–76. Lumley, T. 2005. Assessing Second Language Writing: The Rater’s Perspective. Peter Lang. MacWhinney, B. 2000. The CHILDES Project: Tools for Analyzing Talk. Erlbaum. Malvern, D. and B. Richards. 1997. ‘A new measure of lexical diversity’ in A. Ryan and A. Wray (eds): Evolving Models of Language. Multilingual Matters, pp. 58–71. Malvern, D. and B. Richards. 2002. ‘Investigating accommodation in language proficiency interviews using a new measure of lexical diversity,’ Language Testing 19: 85–104. Malvern, D., B. Richards, N. Chipere, and P. Dura´n. 2004. Lexical Diversity and Language Development: Quantification and Assessment. Palgrave Macmillan. McCarthy, P. M. and S. Jarvis. 2007. ‘Vocd: A theoretical and empirical evaluation,’ Language Testing 24: 459–88. McClure, E. 1991. ‘A comparison of lexical strategies in L1 and L2 written English narratives,’ Pragmatics and Language Learning 2: 141–54. Meara, P. 2005. ‘Lexical frequency profiles: A Monte Carlo analysis,’ Applied Linguistics 26: 32–47. Meara, P. and H. Bell. 2001. ‘P_lex: A simple and effective way of describing the lexical characteristics of short L2 texts,’ Prospect 16: 5–19. Nation, P. 2007. ‘Fundamental issues in modelling and assessing vocabulary knowledge’ in H. Daller, J. Milton, and J. Treffers Daller (eds): Modelling and Assessing Vocabulary Knowledge. Cambridge University Press, pp. 35–43.

G. YU

Wright, H. H., S. W. Silverman, and M. Newhoff. 2003. ‘Measures of lexical diversity in aphasia,’ Aphasiology 17: 443–52. Yule, G. U. 1944. The Statistical Study of Literary Vocabulary. Cambridge University Press. Zareva, A., P. Schwanenflugel, and Y. Nikolova. 2005. ‘Relationship between lexical competence and language proficiency – variable sensitivity,’ Studies in Second Language Acquisition 27: 567–95. Zechner, K., I. I. Bejar, and R. Hemat 2007. Toward an Understanding of the Role of Speech Recognition in Non-Native Speech Assessment (TOEFL iBT research report 2). Educational Testing Service.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

negotiations,’ Group Decision and Negotiation 15: 127–40. Tabachnick, B. G. and L. S. Fidell. 2003. Using Multivariate Statistics. Allyn and Bacon. Tweedie, F. J. and R. H. Baayen. 1998. ‘How variable may a constant be? Measures of lexical richness in perspective,’ Computers and the Humanities 32: 323–52. University of Michigan. 2003. Michigan English Language Assessment Battery Technical Manual 2003. Vermeer, A. 2000. ‘Coming to grips with lexical richness in spontaneous speech data,’ Language Testing 17: 65–83. Wimmer, G. and G. Altmann. 1999. ‘On vocabulary richness,’ Journal of Quantitative Linguistics 6: 1–9.

259

Applied Linguistics: 31/2: 260–281 ß Oxford University Press 2009 doi:10.1093/applin/amp026 Advance Access published on 18 June 2009

Discourse Particles in Corpus Data and Textbooks: The Case of Well PHOENIX W. Y. LAM The Open University of Hong Kong, Hong Kong

INTRODUCTION Discourse particles such as okay, so and well are syntactically optional linguistic items which have no or little propositional value but serve important pragmatic functions. Apart from the term ‘discourse particle’ (see, for example, Aijmer 2002; Fischer 2006), there is no shortage of contenders which are used in similar and often partly overlapping ways in research,1 including ‘discourse marker’ (see, for example, Schiffrin 1987; Schourup 2001) and ‘pragmatic marker’ (see, for example, Watts 1988; Brinton 1996). Because of their importance as ‘sharing devices and intimacy signals in our everyday talk’ (Quirk 1955: 179), they are indispensable in spoken discourse. In the pedagogical setting, however, discourse particles are often dismissed as a sign of dysfluency and their use discouraged (Erman 1987). Although in recent years an increasing number of studies have pointed out that the use of discourse particles contributes to the pragmatic and communicative competence of speakers (Wierzbicka 1991; Mu¨ller 2005) and hence is an essential aspect of language that learners should master, very few have actually looked at the coverage of discourse particles in teaching materials. Previous studies of discourse particles related to language learning, if any, rest on reporting the misuse by learners, judged by the comparison with native data from outside the learners’ community. The pedagogical significance of discourse particles in the foreign language classroom is therefore severely understudied (Fung and Carter 2007). As an

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Discourse particles are ubiquitous in spoken discourse. Yet despite their pervasiveness very few studies attempt to look at their use in the pedagogical setting. Drawing on data from an intercultural corpus of speech and a textbook database, the present study compares the use of discourse particles by expert users of English in Hong Kong with their descriptions and presentations in textbooks designed for learners of English in the same community. Specifically, it investigates the similarities and differences in the use of the discourse particle well between the two datasets in terms of its frequency of occurrence, its positional preference and its discourse function. Results from the analysis show that there are vast differences as regards how the particle well is used in real-world examples and how its use is described and presented in teaching materials. This raises the question to what extent foreign language learners who have minimal exposure to naturally-occurring spoken interactions in English could effectively master the use of discourse particles if they solely rely on these textbooks.

P. W. Y. LAM

261

ROLE OF DISCOURSE PARTICLES IN PEDAGOGY That language in textbooks often differs from language in use has been wellreported in numerous studies. Areas which have been examined are wideranging, with examples such as epistemic modality (Holmes 1988), future time expressions (Mindt 1997), dialogues (Carter 1998; Wong 2002) and multi-word lexical phrases (Koprowski 2005). What seems to be missing, though, is the study of the role of discourse particles in the pedagogical context, especially given the enormous attention these items have received over the years in discourse analysis. As items which occur frequently in speech, discourse particles are characteristic of the spoken language. Their importance in everyday talk can hardly be overstated. Discourse particles facilitate the processes of interpretation and social involvement in spoken interaction (Watts 1988). They are essential to the maintenance of conversational cooperation (Leech and Svartvik 2002). Acting as a discourse lubricant, they ensure interactions go on smoothly. From a pedagogical perspective, failing to master the use of discourse particles may seriously impair the communicative competence of learners (Wierzbicka 1991). Without these items in their speech, learners may come across as unnatural, dogmatic and/or incoherent, hence leading to a greater possibility of communicative failure (Fraser 1990; Brinton 1996). Despite their ubiquity in speech and their importance in discourse and pedagogy, discourse particles are largely overlooked in language learning, with only a few studies exploring their (mis)use by learners of English. Romero Trillo (2002), for example, investigates the development of pragmatic markers including listen, well, and you know in the speech of Spanish speakers of English. His study shows that the non-native adults’ usage of discourse particles is even more limited than that of native children in terms of both

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

attempt to bridge this gap, the present study looks at how the discourse particle well is used by expert speakers in an intercultural spoken corpus in Hong Kong and how it is described and presented in textbooks designed for learners within the same community. In particular, the frequency of occurrence, positional preference, and discourse functions of the particle in the two data sources will be compared both qualitatively and quantitatively to investigate the extent to which textbook data reflects real-world usage. A typical example of discourse particles, well is one of the most frequently occurring items and probably the most thoroughly researched (Schourup 2001; Aijmer 2002). Its high frequencies of occurrence in the two data sources enable a meaningful quantitative analysis to be conducted. Its versatility in serving different pragmatic functions in discourse has been well-documented in previous studies (see, for example, Lakoff 1973; Schiffrin 1987). Consequently, its pragmatic functions may have been better understood than others, which in turn allows for a more reliable comparison to be made between its actual use and textbook descriptions.

262 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

quantity and diversity. Such a shortage of discourse particles in the speech of the non-native speakers, he argues, could make their speech sound blunt and impolite. Similarly, Mu¨ller (2004, 2005) reports differences in the use of discourse markers by German EFL speakers and American native speakers in a movie-telling experiment. Non-native data, for instance, exhibit fewer functions of well when compared with native data. Her findings point to the possible influence of textbooks in the use of discourse particles by German learners, though only a cursory examination of the items well and so in textbooks based on the ratio of occurrence of the two elements is presented. The use of pragmatic markers in Xhosa English, a sub-variety of Black African English, is the focus of de Klerk (2005). Differences found in the use of well by Xhosa speakers of English and native speakers again are linked to the observation that discourse particles are largely overlooked in the local educational system, though no empirical evidence is given. Recently, Fung and Carter (2007) have investigated the production of discourse markers in pedagogic settings using data of British native speakers and Hong Kong learners of English. Similar to the findings presented in Romero Trillo (2002), Fung and Carter (2007) report that native speakers use a wider range of discourse markers including actually, right, and well to achieve a broader variety of pragmatic functions such as marking responses and attitudes. This leads to their conclusion that the misuse of discourse particles can be a communication obstacle for learners. While the comparative studies discussed above offer valuable insights into the differences in the use of discourse particles by native and non-native speakers of English, they only hint at some possible pedagogical implications. Specifically, whether and how discourse particles are taught remains an unsettled issue. Admittedly, the study of the production of discourse particles by language learners could help us identify the potential problems they have in using these items. Equally important at the same time, however, is the issue of how discourse particles are presented in teaching materials and whether these learning resources genuinely mirror naturally-occurring language. As pointed out in many studies, textbooks play a central role in syllabus design and lesson planning (Olson 1980; Kramsch 1988). For EFL learners who do not always have the opportunities to immerse in an English-speaking environment outside the foreign language classroom, textbooks constitute the bread and butter of their language learning experience. In the context of Hong Kong, the centrality of textbooks can be best summarized by Reynolds’ (1974: 41) observation: ‘the textbook is taught, not the students’. The role of textbooks is therefore of particular importance when it comes to supplying learners with the right information as regards one of the most common features in spoken English. Another important issue which is raised by the above review concerns a popular comparative approach used in previous studies, namely the dichotomy between native and non-native speakers of English. In recent years, however, the very concept of native speakers of a language has been increasingly

P. W. Y. LAM

263

DATA AND METHOD In the present study, data are drawn from two sources, namely the Hong Kong Corpus of Spoken English (HKCSE thereafter) and a textbook database. The HKCSE is an intercultural corpus containing approximately 1 million words of naturally-occurring speech (see Cheng et al. 2005, for further details of the corpus). Compiled between the mid-1990s and the turn of the millennium, the corpus consists of 311 texts which are primarily intercultural encounters in English between Hong Kong Chinese whose first language is Cantonese and speakers of languages other than Cantonese, though in a few settings only Hong Kong Chinese are involved. As the main purpose of building the HKCSE is to investigate the linguistic features of spoken English produced by successful users of English in Hong Kong in a variety of academic, business, social and professional settings, participants in the corpus are all competent

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

challenged (Davies 2003, 2004), particularly following the rise of the global role of English (Graddol 1997, 2006). The prevalence of intercultural encounters in English has also generated the heated debate on the ownership of the language. In the field of English language teaching, there have been rising reservations about whether EFL learners should follow native English models as ‘standards’ given the possible historical, cultural and contextual differences concerned (see, for example, Kachru 1996; Seidlhofer 2002), not to mention the implications of language imperialism and cultural hegemony (Phillipson 1992; Pennycook 1998). In order to compare the actual performance of competent speakers with teaching materials used in the same community, the present study makes use of data from an intercultural corpus of speech of ‘successful users of English’ (Prodromou 2003) in the context of Hong Kong. As interlocutors in the corpus are all effective communicators of English regardless of their first language, they reflect more faithfully the actual usage and thus language needs of the local community than native speakers from elsewhere do. As Bhatt (2005: 48, emphasis mine) rightly puts it, ‘standard language has to be treated as endonormatively evolving from within each community according to its own histories and cultures of usage. Standards can’t be imposed exonormatively from outside one community’. A comparison of Hong Kong textbook data with naturally-occurring examples from the same community thus is deemed more suitable as it relevantly and realistically contrasts local uses and needs of language with local training and practice as provided by textbooks. This does not imply, however, that findings from the present study are only pertinent to the restricted context of Hong Kong. Rather, the present study contributes to the important discussion of the appropriate language models for EFL learners and illustrates how a spoken corpus of intercultural encounters can be exploited to investigate the natural usage of particles in an ever-increasing number of intercultural communities around the globe, where the language backgrounds of speakers are becoming more and more diverse and complex.

264 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

speakers of English who regularly and successfully communicate in English either professionally or socially. Through years of formal education and extensive interactions with other expert speakers, they have come to represent a cross-section of the adult population in Hong Kong who obtain a high proficiency in English regardless of their mother tongue. As far as the communicative competence in English in the context of Hong Kong is concerned, they probably better signify what school students may aspire to become than native speakers from an external community. To provide a description of discourse particles in teaching materials which is lacking in the literature and to compare particle usage in ‘real’ English and ‘school’ English, a database consisting of English textbooks collected in Hong Kong was created. Textbooks designed for upper secondary students in Hong Kong were chosen because teaching materials at that level have a strong focus on the spoken language. This is largely owing to the popular demand from the local market to cater for the preparation of the university-entry public oral examination which consists of two assessment tasks, namely an individual presentation and a group discussion with other candidates. Presumably, these coursebooks should have a more extensive coverage of oral features such as discourse particles than textbooks at other levels. Altogether 15 textbooks which represent a sample of mainstream English coursebooks designed specifically for upper-secondary students in Hong Kong were collected and analysed.2 Produced by seven publishers from the year 1994 to 2003, the coursebooks selected are used for the 2-year long upper-secondary study in Hong Kong. Two-thirds are specifically designed for the training of oral skills while the remaining are general textbooks which cover the four skills of listening, reading, speaking and writing. In the general textbooks, only the speaking section is examined. To facilitate comparisons with the HKCSE and among teaching materials, all textbooks in the database contain some teaching points or explanations. Drill exercise books which only consist of gap-filling practice materials or exam papers are not included. Previous studies of teaching materials have provided very few guiding principles as regards what should be included in a textbook database (Nelson 2000), especially when quantitative analysis is conducted. To give a comprehensive account of well in the textbooks examined, the present study provides a qualitative and quantitative account of the descriptions and actual usage of the particle in context in textbooks. Accordingly, the contents of the textbooks are categorized into two parts: the teaching section and the sample section. The teaching section contains teaching points, explanations and short examples of language which are suggested to be used in presentations and discussions. These examples are only studied qualitatively to see if textbook descriptions of well mirror usage in corpus data. The sample section contains ‘model’ presentations and discussions which are suggested to be exemplary texts for the local oral examination. These sample texts resemble the basic structures of naturally-occurring presentations and discussions and thus are analysed both qualitatively and quantitatively just as texts in the HKCSE. This draws

P. W. Y. LAM

265

a direct comparison of the use of well between invented texts in textbooks and naturally-occurring data in the HKCSE. While the present study is concerned with the use of well as a discourse particle (D-use), the propositional use of the word (P-use) will also be briefly discussed to demonstrate how typically the word achieves discourse functions.

FINDINGS

Frequency of occurrence One of the most striking differences between naturally-occurring data in the HKCSE and sample texts in the textbook database in relation to the discourse use of well is the rate at which the particle appears in discourse. Table 1 compares the use of well in the two datasets under investigation. A number of observations can be made from Table 1. First, there is a considerable difference in terms of the total number of well in the two sample text types in textbooks. While the total number of words in textbook discussions (n = 13,032) is only 20% more than that in textbook presentations (n = 10,817), the number of well in discussion texts is twelve times more than that in presentations (n = 117 versus n = 9). This indicates that the word occurs much more frequently in textbook discussions. In addition, there is a huge discrepancy in the discourse-function ratio of well (D-use/total use as

Table 1: The comparison of well in the textbook sample section and in the HKCSE

Total number of words Total number of well Number of D-use Number of P-use Unclassified use D-use/total use (%) D-rate (per 10,000 words)

Textbook presentations (N = 38)

Textbook discussions (N = 11)

HKCSE (N = 311)

10,817 9 1 8 0 11.11 0.92

13,032 117 107 10 0 91.45 82.11

949,972 2,714 1,913 796 5 67.99 19.56

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

This section examines and compares the use of well in the HKCSE and the textbook database. In particular, the analysis focuses on three aspects of D-use well, namely its frequency of occurrence, position and discourse function. Because of the nature of the data, quantitative analyses on these three aspects are not carried out in the textbook teaching section.

266 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

Table 2: The comparison of the discourse rate of well in presentations in the HKCSE and in the textbook sample section Data source

Text type

Discourse rate of well (per 10,000 words)

HKCSE

Business presentation Lecture Speech Sample presentation

5.47 12.20 5.07 0.92

Textbook database

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

expressed in per cent) between the two sample text types. While only 11.11% of instances of well are used as a discourse particle in textbook presentations, the percentage rises to 91.45% in textbook discussion texts. This shows that when the word is used it is most likely to be a particle in textbook discussions and vice versa in presentations. In consequence, the textbooks give a strong impression that well as a discourse particle is highly common in discussions but almost nonexistent in presentations. Such a noticeable contrast in the number of D-use well in textbook presentations (n = 1) and discussions (n = 107) is reflected in the discourse rate of well in the two text types. The discourse rate here refers to the number of discourse use of well in a sample of 10,000 words. In a 10,000-word sample of textbook presentations, the discourse rate of well is merely 0.92. In textbook discussions, on the other hand, the discourse rate is 82.11. Again, this signifies that the discourse use of well occurs considerably more frequently in textbook discussions than in textbook presentations. When compared with corpus data, the use of well in textbooks also shows highly contrastive patterns. As regards the discourse–function ratio, textbook discussions have the highest figure among the three data sources while the ratio in textbook presentation texts is unusually low. A similar pattern can be observed concerning the discourse rate of well, with the highest rate found in textbook discussion texts and the lowest in textbook presentations. These findings appear to suggest an irregularity in the discourse rate of well in textbooks when compared with naturally-occurring data. This is substantiated by a further comparison between textbook texts and their corresponding text types in the corpus. Given the rather formal and overwhelmingly monologic nature of sample presentations in textbooks, they are compared with business presentations, lectures and speeches in the HKCSE. For textbook discussions which involve a clear agenda and the negotiation of ideas, business interviews, meetings and public TV talk shows in the corpus are chosen for contrast. The comparison of the discourse rate of well in some major monologic texts in the HKCSE and in textbook presentations is shown in Table 2.

P. W. Y. LAM

267

Position In the discussion of the position of discourse particles, various units of talk have been considered, including turn, utterance, tone unit, sentence and clause-element (see, for example, Stenstro¨m 1990; Jucker and Ziv 1998; Schourup 1999). The term utterance is used in this study as it probably better accommodates features such as fillers, incomplete structures and overlapping speech which are common in spontaneous spoken discourse than notions such as sentences and clauses which are more geared towards

Table 3: The comparison of the discourse rate of well in discussions in the HKCSE and in the textbook sample section Data source

Text type

Discourse rate of well (per 10,000 words)

HKCSE

Business interview Business meeting Public TV talk show Sample discussion

22.13 30.33 34.65 82.11

Textbook database

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Table 2 illustrates a marked difference between textbook and naturallyoccurring presentations in terms of the discourse rate of well. In heavily scripted business presentations and public speeches in the HKCSE, the discourse rate of well is at least five times more than that in textbook presentations. The difference is even more apparent if textbook presentations are compared with lectures in the corpus, which are more spontaneous in nature. If these textbook presentations are designed to teach students the features of spoken language in contexts which they are most likely to encounter in future, one may wonder why textbook presentations have such a low discourse rate of well. Interestingly, the contrast in the discourse rate of well between textbook and naturally-occurring discussions shows exactly opposite results. Table 3 compares the discourse rate of well in some major discussion texts in the HKCSE and in textbook discussions. Table 3 shows that the sample discussions in textbooks have a much higher discourse rate of well when compared with discussion texts in naturallyoccurring speech. The rate in textbooks (82.11 per 10,000 words) is at least twice the rates in the HKCSE, which only range from 22.13 to 34.65. This suggests that there is an unusually high rate of D-use well in textbook discussions when compared with similar texts in corpus data. Again, it remains doubtful why the particle well should be used so frequently in these textbook discussions if they reflect natural usage.

268 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

the written language than the spoken language. Following Stenstro¨m (1994: 226), an utterance is defined as ‘anything that a speaker says’. Table 4 compares the distribution of utterance position of D-use well in the HKCSE and the textbook sample section. As a discourse particle, well mostly occurs in utterance initial and medial positions in the HKCSE. Altogether tokens in these two positions constitute 96% of the total number of discourse use of well. Specifically, more than half of the discourse tokens of well (57.9%) are utterance initial, with a slightly lower proportion of occurrences (38.1%) embedded in discourse. The use of well as a discourse particle in utterance initial position is illustrated in Example 1:3

Example 2 shows an instance of well in medial position serving discourse function: (2) . . . so let’s talk about our flow well Hong Kong is . . . (HKCSE, B154)

The textbooks examined, on the other hand, show a different positional distribution of the particle. Of the 108 instances of D-use well found in textbook presentations and discussions, 97 instances occur at the beginning of an utterance, making up 89.8% of the total number of pragmatic use. Example 3 shows an utterance initial well in a textbook discussion: (3) Well, look, let’s try to summarize what’s been said so far . . . (Potter 2003a: 129)

In contrast, only eleven tokens (10.2%) are found medially, making them a small minority in the textbook database. An utterance medial D-use well is

Table 4: The distribution of utterance position of D-use well in the HKCSE and in the textbook sample section Position

D-use well in HKCSE

Initial Medial Final Stand-alone

1,107 728 14 64

Total

1,913 (100%)

(57.9%) (38.1%) (0.7%) (3.3%)

D-use well in sample texts in textbook database 97 11 0 0

(89.8%) (10.2%) (0.0%) (0.0%)

108 (100%)

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(1) Well you’re really busy . . . (HKCSE, A041)

P. W. Y. LAM

269

shown in Example 4: (4) O.K. What’s next? An airport . . . well, they’ll be coming by plane, so they’ll see the new airport when they arrive . . . (Sutton and Duncan 1999: 154)

In addition, no instances of D-use well are found in final and stand-alone positions in textbooks, while a small number can be found in the corpus. Example 5 shows the use of well as a single-word utterance from the HKCSE:

As regards the main positions in which D-use well occupies, figures from Table 4 clearly indicate that there is a much higher proportion of initial Duse well and a much lower proportion of medial D-use well in textbooks when compared with corpus data. The high percentage of well in initial position in textbooks seems to suggest that initial D-use well is vastly dominant, which unfortunately is not validated by the corpus findings.

Discourse function This section first discusses the discourse functions of well as identified from the HKCSE. This is followed by a qualitative analysis of how the discourse functions of well are described in the teaching section of the textbooks. Finally, a quantitative comparison of the functional distribution of D-use well in the HKCSE and the textbook sample section is made. In conducting a functional analysis of well, the present study follows a bottom-up corpus-linguistic approach. In other words, the functional taxonomy developed in the analysis is derived from the recurrent patterns observed in the datasets but not from a pre-existing framework or theory. This involves rounds of modification before a classification scheme is devised to fully capture the range of functions observed in the data. In total, 1,913 uses of well as a discourse particle are examined in the HKCSE, of which 1,889 instances are functionally categorized. 24 tokens are functionally unclassified owing to insufficient contextual information. Six major functions of well are identified, which can be categorized into three different domains: textual, interpersonal and interactional. Instances of well expressing functions in the textual domain are concerned with the structuring and organization of discourse. They largely correspond to the text-oriented organizational unit (OT) in Linear Unit Grammar (Sinclair and Mauranen 2006). In Halliday’s (1985) terms, they achieve the textual meta-function. Interpersonal functions are related to the expression of attitudes, emotions, and personal evaluations whereas interactional functions facilitate processes such as planning and turn management

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(5) S1: mm Chung Kai does that satisfy you S2: well S1: probably not entirely I suspect ((laugh)) (HKCSE, P122)

270 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

in the interaction. Tokens of well expressing functions in the interpersonal and interactional domains in the present study realize Halliday’s (1985) interpersonal meta-function. They largely correspond to the interactive-oriented organizational unit (OI) in Linear Unit Grammar (Sinclair and Mauranen 2006). Admittedly, the classification of function is not always a straightforward issue. After all, ‘natural language is not a sharp instrument with absolute or rigid boundaries, but is blurred at the edges’ (Sinclair and Mauranen 2006: 61). Yet given sufficient linguistic and contextual details, it is usually possible to identify the key function in a single instance (Holmes 1984). The discourse functions of well in the HKCSE are discussed in detail as follows.

In the textual functional domain, well serves as a frame (Sinclair and Coulthard 1975) and a link. As noted in the Cambridge Grammar of English, these textual uses of well ‘organize and monitor an ongoing discourse’ (Carter and McCarthy 2006: 901). The framing use of well to insert a point of division or transition for easy comprehension is one of the most frequently occurring functions in the corpus data. In these examples, well acts as a boundary marker in discourse to signal transitions in topic and discourse stage. At times, it plays a role similar to punctuation marks in the written language in dividing words into clauses and sentences. In Example 6, the speaker uses well together with a metalinguistic comment let’s talk a little bit about conflict to indicate a topic change: (6) . . . but yet at the same time not violate our group harmony (.) yea it can be done (.) it can be done okay well let’s talk a little bit about conflict why is conflict management so important . . . (HKCSE, B123)

Apart from segmenting texts, well could be used as a link to introduce explanations and additional information to the preceding discourse. In their work Cohesion in English, Halliday and Hasan (1976: 269) briefly describe this use of well as an introducer of ‘an explanatory comment’. Similarly, the use of well to continue an opinion or an answer is found in Mu¨ller (2005). Example 7 shows a speaker providing additional information about her supervisor following an utterance medial well: (7) . . . so I start er to er write a proposal and I talk to my er supervisor er well he’s a lecturer . . . (HKCSE, B082)

Interpersonal function In the interpersonal domain, the responsive use of well is the most dominant among all the functions identified. This is in accordance with the observation

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Textual function

P. W. Y. LAM

271

made in previous studies that well is commonly known as a marker of response (Lakoff 1973; Schiffrin 1987). As a responsive signal, well is most frequently associated with dispreferred responses such as disagreements and criticisms (Lam 2006), though it is also found to preface direct answers and follow-up responses in the HKCSE and other studies (see, for example, Schourup 2001; Aijmer and Simon-Vandenbergen 2003; Mu¨ller 2005; Lam 2008). Example 8 illustrates the typical use of responsive well in a qualified answer. In combination with the phrase in many cases, it indicates that the speaker’s agreement is only partial:

Another interpersonal function of well is the expression of feelings. In the HKCSE, well is used to convey a range of feelings such as surprise, concession and dismissal (cf. Carlson 1984; Carter and McCarthy 2006). Schourup (2001: 1043) describes this use of well as a ‘mental state interjection’, which shares emotive properties with items such as ouch and wow. As a feeling carrier, well is most commonly found in the HKCSE to convey a sense of resignation. In Example 9, speaker S2 is talking to speaker S1 about her wedding night, which fell within the tournament of the World Cup in 1990. As a football fan, her husband told her that he would spend the whole wedding night watching the World Cup matches. Speaker S1 first indicates his surprise at the husband’s decision with two instances of oh dear. This is followed by a brief pause and repeated uses of well to show a sense of helplessness and resignation: (9) S1: S2: S1: S2: S1: S1: S2: S1:

((laugh)) oh dear and he actually watched the world cup he watched it ((laugh)) oh dear ((laugh)) a true a true football fan there but a bad husband really  mm ((laugh)) yeah oh dear (.) well well well but er I take it he er you know it’s good it’s good job the world cup didn’t last er very long right it comes around every four years so it’s alright . . . (HKCSE, C001)

Interactional function In the interactional domain, well is used for processing purposes and turn management. This is in accordance with corpus findings from the Longman Grammar of Spoken and Written English, which suggest that discourse markers ‘facilitate the ongoing interaction’ (Biber et al. 1999: 140). As a processing

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(8) S1: so the separate rule you are talking about is actually a lower standard S2: well in many cases yes er let me show you an example here . . . (HKCSE, P102)

272 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

device, well allows speakers to signal to the other participants that some internal processing is going on and enables them to gain extra time for their turn. Therefore, it is particularly useful for marking self repair and hesitation (cf. Carter and McCarthy 2006). In Example 10, the speaker encounters a word recovery problem when she talks about the element to which derivational morphemes are added. Notice her use of the filler um and the particle well as place-holders while she undergoes some internal contemplation before resting on her choice of the expression word stem among other options:

Finally, well is used in turn management to signal speakers’ desire to take control of the conversational floor either by floor-holding or turn-taking. Example 11 shows the attempt by the listener (S2) to take the conversational floor from the current speaker (S1) with the use of well when the speaker is in the middle of his talk: (11) S1: er it’s between double and three times more  S2: well you know if you check with your standards (.) we could set the same pay . . . (HKCSE, B148)

In the teaching section of the textbook database, the focus of the teaching materials is often on how to communicate effectively in presentations and discussions, as these are the scenarios students have to face in the local oral examination. Since these textbooks are mainly designed to achieve the goal above, they are repeatedly filled with examples of speech functions which are common in presentations and discussions. Example 12, for instance, shows how the speech function ‘disagreeing with a suggestion’ is claimed to be expressed by the following utterance (Sutton and Duncan 1999: 64): (12) Well, I don’t think that’s possible, because . . . (Sutton and Duncan 1999: 64)

In other words, no paragraphs or sections separately discuss the discourse functions of well. The various uses of the particle can only be found in examples showing how different speech functions are realized in the textbooks. These examples are all short, detached instances of language with minimal provision of contextual information such as Example 12. The functional determination of these examples of well is therefore largely based on the speech function suggested by the textbook authors. An analysis of these instances shows that all speech functions associated with well in the textbook teaching section could be subsumed under the functional categories identified in the HKCSE.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(10) . . . but what I’m saying is (.) derivational morphemes they usually when they add to um well when they add to a word stem they entail a change of meaning . . . (HKCSE, A028)

P. W. Y. LAM

273

In the textual domain, well is found in examples related to the maintenance of the overall flow of presentations and discussions. This includes initiating a topic in a presentation, starting and ending a discussion and moving on to the next point in a discussion. Example 13 illustrates the use of well as an initiator in a textbook discussion text: (13) Well, shall I start? I suggest we divide our discussion into three parts . . . (Leetch 2002: 62)

(14) Well, I really don’t agree at all! (Sutton and Duncan 1999: 56)

As mentioned before, well is frequently associated with dispreferred responses in the corpus. However, examples like the one above are not found in the HKCSE. In fact, a cursory search reveals that the word combination well I really don’t agree at all is not present in the corpus, nor does it seem to be a common way to express disagreements in corpora of other varieties of English. While creativity in language should not be discouraged, in this particular context it is perhaps more helpful to exclude rare linguistic expressions, for they may run the risk of misleading learners by providing examples of language which are hardly used in natural interactions (Holmes 1988; Gilmore 2004; Koprowski 2005). As far as the interactional aspect of well is concerned, the uses of well as a processing device and as a turn managing signal are both found in the teaching section. In one of the textbooks examined, it is suggested that the particle is one of the ‘hesitation words’ along a wide range of lexical items such as er, in fact and let me see which students could employ when they need time for

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

These examples in the textbooks highlight the discourse use of well in topic and discourse stage management and largely correspond to the frame function identified in the corpus. The use of well as a linking device, however, is not discussed in the teaching section of the textbook database. As regards the interpersonal aspect of well, no instances of D-use well in the textbook teaching section are found to be associated with affective meaning. The responsive function of well, however, can be readily observed in examples illustrating how disagreements and qualified agreements should be expressed. In one of the textbooks where well receives a brief mention, it is suggested that well is one of the lexical items used to ‘soften disagreement and doubt’ (Potter 2003a: 42). However, it is not obvious how the word is used to achieve this function as no examples are given in the textbook. Other kinds of dispreferred responses associated with well in the teaching section include agreeing but not in a very enthusiastic way (Esser 1999) and ‘avoiding giving an opinion’ (Sutton and Duncan 1999: 56). In addition, well is found as part of a textbook example which serves to illustrate ‘disagreeing strongly with someone’ (ibid.), reproduced here as Example 14:

274 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

planning what to say (Potter 2003a: 25). With regard to turn management, well is associated with the speech function ‘interrupting politely’ (Potter 2003b: 39). This corresponds to the turn-taking sub-type identified in the corpus. Again, examples given in the textbook teaching section are decontextualized and it remains unclear how the particle assists in politely taking the conversational floor, as the preceding utterance is not provided. A case in point is Example 15, which is given in a textbook to illustrate how to interrupt without being rude:

When compared with the textbook teaching section, the sample section generally provides more contextual information as the sample presentations and discussions have structures resembling naturally-occurring texts of a similar nature. Consequently, the functions of the particles can be examined together with the linguistic co-texts in the samples. Table 5 compares the functional distribution of D-use well in the textbook sample section and the HKCSE. As discussed in the previous section regarding the frequency of occurrence of well in the textbooks, only a single instance of D-use well is found in sample presentations, acting as a frame in the text. For sample discussions, the use of well in responses constitutes 74 out of a total of 107 instances. It is therefore the most dominant discourse function in textbook discussions, making up more than two-thirds (69.2%) of the total. The framing use of well contributes another 24.3%. The other four functions merely make up a small proportion (6.5%) of use altogether. In the HKCSE, the responsive and framing uses of well are most common. Each constitutes roughly one-third of the total use. The remaining proportion (29.6%) is shared between the other four functions, with the processing function being noticeably more dominant (17.7%) than the other three. Table 5 thus essentially shows that there are major discrepancies concerning the functional distribution of well in real-world situations and in teaching materials. While it is true that the framing and responsive uses are the two key pragmatic functions found in the corpus, the other four functions also add up to a reasonable proportion in the data. In the textbooks, however, these four functions only contribute a very small number of instances. In addition, the responsive function of well appears to be over-emphasized while the frame function is slightly overlooked in textbooks. When compared with findings from the HKCSE, textbook discussions contain a considerably higher (69.2% versus 37.5%) proportion of responsive well and a relatively lower proportion of well as a frame (24.3% versus 32.9%). Given the proportions of framing and responsive well in sample discussions and the negligible number of well in sample presentations for textual organization, textbook writers seem to pay excessive attention to well in responses while ignoring other important

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

(15) Well, I think the solution is simple. We should . . . (Potter 2003b: 39)

HKCSE

Sample discussion texts

Sample presentation texts

Count Percentage within function Count Percentage within function Count Percentage within function 621 32.9%

26 24.3%

1 100.0%

Framing

Function

70 3.7%

1 0.9%

0 0.0%

Linking

708 37.5%

74 69.2%

0 0.0%

Responsive

70 3.7%

2 1.9%

0 0.0%

Emotive

Turn managing 0 0.0% 1 0.9% 85 4.5%

Processing 0 0.0% 3 2.8% 335 17.7%

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Source of data

Table 5: The functional distribution of well in the textbook sample section and in the HKCSE

1889 100.0%

107 100.0%

1 100.0%

Total

P. W. Y. LAM 275

276 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

discourse functions such as framing which are also highly common in naturally-occurring spoken examples.

DISCUSSION

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Despite the fact that well is one of the most frequently occurring words in the spoken language and many of the 15 textbooks collected for the present study claim to focus on oral skills, none of them devotes a separate section or paragraph to the description of the particle, not to mention the discussion of discourse particles as a collective group. Given the importance of well to express various textual, interpersonal and interactional functions in talk as supported by the corpus findings, such treatments seem to be less than adequate. The analysis of well in the HKCSE and the textbooks above demonstrates that noticeable differences exist in terms of the frequency of occurrence, position and function of well between textbook data and naturally-occurring speech. The vast difference in the discourse rates of well between textbook presentations and discussions gives a strong feeling that the particle is virtually nonexistent in presentations while ubiquitous in discussions. This, however, is not substantiated by the corpus evidence in the present study. While most of the instances of well found in the textbook sample section occur in utterance initial position, corpus evidence suggests that the particle is also fairly common in medial position. In addition, the functions of well as described and realized in textbooks do not seem to be a close match with their functions in corpus data. While it is understandable that minor pragmatic functions such as the linking and emotive uses of well are omitted in textbooks, it remains unclear why such a dominant function as framing is absent in presentations, a monologic text type in which the organization and structuring of discourse are essential. Furthermore, the responsive use of well appears to be over-stressed in sample texts. This largely arises from the observation that expressing disagreements appropriately in discussions is considered an important teaching goal by the textbook writers and that well appears to be almost obligatory in such examples. Although the responsive use of the particle is also typical in corpus data, an over-emphasis on this particular function in textbooks could possibly lead to an overuse of well in responses by students at the expense of overlooking other important functions. Of course, a mismatch between textbook data and corpus data does not necessarily mean that such teaching materials are pedagogically inappropriate. After all, authenticity is not the only, if ever a primary, criterion for textbook design, not to mention the controversy surrounding the term ‘authentic’ (for the discussion of authenticity in language teaching and learning, see Breen 1985; Cook 2001; Widdowson 2003). As admitted by Carter (1998), ‘real’ English could be more difficult to understand and produce. Therefore, it may not always be pedagogically the most practical and effective, especially if materials designed for beginners are concerned. Even if such is the case, a corpus is only a glimpse of a given language variety at a restricted period

P. W. Y. LAM

277

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

in time (McEnery and Wilson 1996) and there are aspects of language which may not be fully unveiled by corpus evidence (Cook 1998). Yet for a crucial pragmatic resource like discourse particles which are almost impossible to miss in any contemporary spoken corpus, it is hard to justify why textbook writers fail to give a more detailed and accurate description and presentation of how they are used in real-life situations. Considering the fact that many of the textbooks examined are specially designed for training English oral skills for intermediate to advanced learners in Hong Kong in particular, such a mismatch seems to be at odds with the aim of the oral examination that these textbooks cater for, which is ‘to test the ability of candidates to understand and use spoken English for practical communication as it might be encountered in academic or vocational situations’ (Hong Kong Examinations and Assessment Authority 2005: 3). One might therefore, quite reasonably, query whether such textbooks are so detached from reality that they have ultimately lost their pedagogical value. In recent years, it has been increasingly acknowledged that discourse particles are crucial for learners to communicate successfully at the pragmatic level of interaction. If language learners are denied access to these critical pragmatic devices in their learning process, they may not be able to fully project their personality in the target language. As a result, although they may well be capable of attaining transactional goals, they could only operate in the target language with a ‘reduced’ personality (McCarthy 1998: 112). In other words, learners are deprived of the right to behave and express themselves in the same way as they do in their mother tongue. The image that they could present in the second or foreign language is, at most, a partial alter ego. At worst, the dearth of discourse particles in their talk could leave them ‘potentially disempowered and at risk of becoming a second-class participant’ (O’Keeffe et al. 2007: 39). Given that discourse particles are crucial in achieving pragmatic competence and that the descriptions and examples of discourse particles in the textbooks examined are far from satisfactory, substantial revisions with the incorporation of naturally-occurring examples are required in order to present a more comprehensive picture to students concerning how discourse particles are used. In this connection, numerous practical suggestions concerning how language corpora can be used to enhance teaching and learning activities have been made. In order to make such corpus findings and applications more accessible to the pedagogical setting, closer ties with the teaching profession can be established through educational conferences and collaboration with schools. O’Keeffe and Farr (2003) take a step forward and argue for the integration of corpus linguistics into teacher education courses. Furthermore, corpus linguists have taken an active role as materials writers to introduce corpus evidence into coursebooks (see, for example, Carter et al. 2000, for the teaching of discourse particles using authentic examples, and McCarthy et al. 2005a,b, for a corpus-informed syllabus of conversational strategies). At the same time, the availability of many mega-corpora in the public domain means that

278 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

CONCLUSION Through the unique examination of the particle well in an intercultural corpus of spoken English in Hong Kong and its descriptions and presentations in local textbooks, the present study has provided an example of how language is taught and how language is actually used in the same community. Findings from the present study have shown that wide discrepancies are found between teaching materials and naturally-occurring examples in one of the most frequently used discourse particles in the English language. This raises the issue to what extent these textbooks examined reflect natural usage and hence allow learners to be aware of how discourse particles are used. This article argues that discourse particles are a valuable linguistic resource which learners have a right to gain access to. Textbooks therefore should at least describe and present them in the ways they are used in naturally-occurring examples so that students can have a basic understanding of these items. Whether students want to exploit this resource for productive use or only for reception purposes is at their discretion (cf. Fung and Carter 2007), though they should be given the choice all the same, especially for language learners at a more advanced stage. When the issue of critical language awareness has become part

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

textbook writers now have at their disposal easy access to a large quantity of naturally-occurring texts to confirm or refute their intuitions. Major publishing houses even have their own in-house corpora on which teaching materials can be based. In the cases where sizeable corpora from a specific EFL community are not available to materials designers, corpus-based dictionaries and grammars can be consulted (see, for example, Biber et al. 1999; Carter and McCarthy 2006), with the careful supplement and comparison of data and studies relevant to the local setting. Findings from the present study which are generated from a large number of naturally-occurring examples in a wide range of settings in Hong Kong, for example, provide an empirical basis for the improvement of the descriptions of discourse particles in local textbooks as well as in grammars and dictionaries of world Englishes. With the increasing number of research outputs showing the fruitful results of datadriven learning (DDL) (see, for example, Tribble and Jones 1990; Johns 1991), it is high time to apply the methods of corpus linguistics to language teaching and learning, especially in areas such as discourse particles, where invented decontextualized examples could hardly elucidate to students the many discourse functions of these linguistic items in a variety of contextual situations. In this respect, a corpus-based and data-driven approach to learning discourse particles with the use of concordancers may be useful (see, for example Zorzi 2001, on the learning of Italian discourse markers and Mo¨llering 2004, on the teaching of German modal particles) and could probably provide the right resources for learners, especially for intermediate to advanced learners, to explore and research in a more learner-centred way how discourse particles are actually used.

P. W. Y. LAM

279

of the agenda for language learning, it is perhaps also appropriate to review the place of discourse particles in language learning.

SUPPLEMENTARY DATA Supplementary material is available at Applied Linguistics online.

ACKNOWLEDGEMENTS

NOTES 1

2

In the literature, there is a lack of a universally accepted term which most people working in the field give consent to. For detailed discussions, see Brinton (1996) and Fischer (2006). For the sake of brevity, details of the textbooks used in this study are given in the supplementary appendix, which

3

is available in the online version of the article. Notes on transcription: ‘S1/S2:’ example in the HKCSE with more than one speaker; ‘. . .’: an utterance reported only in part; ‘’: the beginning of overlapping talk; ‘(.)’: an unfilled pause.

REFERENCES Aijmer, K. 2002. English Discourse Particles: Evidence from a Corpus. John Benjamins. Aijmer, K. and A.-M. Simon-Vandenbergen. 2003. ‘The discourse particle well and its equivalents in Swedish and Dutch,’ Linguistics 41/6: 1123–61. Bhatt, R. M. 2005. ‘Expert discourses, local practices and hybridity: The case of Indian Englishes’ in A. S. Canagarajah (ed.): Reclaiming the Local in Language Policy and Practice. Lawrence Erlbaum. Biber, D., S. Johansson, G. Leech, S. Conrad, and E. Finegan. 1999. Longman Grammar of Spoken and Written English. Longman. Breen, M. 1985. ‘Authenticity in the language classroom,’ Applied Linguistics 6: 60–70.

Brinton, L. J. 1996. Pragmatic Markers in English: Grammaticalization and Discourse Functions. Mouton de Gruyter. Carlson, L. 1984. ‘Well’ in Dialogue Games: A Discourse Analysis of the Interjection ‘Well’ in Idealized Conversation. John Benjamins. Carter, R. 1998. ‘Orders of reality: CANCODE, communication, and culture,’ ELT Journal 52/1: 43–56. Carter, R., R. Hughes, and M. McCarthy. 2000. Exploring Grammar in Context. Upper-intermediate and Advanced. Cambridge University Press. Carter, R. and M. McCarthy. 2006. Cambridge Grammar of English. Cambridge University Press.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

I would like to thank the editor of the journal and three anonymous reviewers for their valuable comments on an earlier version of this article. The bulk of the work described in this article was carried out at The Hong Kong Polytechnic University and was substantially supported by grants from the Research Grants Council of the Hong Kong Special Administrative Region (Project No. B-Q714). I am deeply indebted to Prof. Winnie Cheng and Prof. Martin Warren for their generous permission to let me use the Hong Kong Corpus of Spoken English prior to its completion and publication.

280 DISCOURSE PARTICLES IN CORPUS DATA AND TEXTBOOKS

Hong Kong Examinations and Assessment Authority. Johns, T. 1991. ‘Should you be persuaded – two samples of data driven learning materials,’ English Language Research Journal 4: 1–16. Jucker, A. H. and Y. Ziv. 1998. ‘Discourse markets: Introduction’ in A. H. Jucker and Y. Ziv (eds): Discourse Markers: Descriptions and Theory. John Benjamins. Kachru, B. 1996. ‘English as lingua franca’ in H. Goebl, P. Nelde, Z. Zary, and W. Wo¨lck (eds): Kontaktlinguistik. Contact Linguistics. Linguistique de Contact. Vol. 1. Mouton de Gryuter. Koprowski, M. 2005. ‘Investigating the usefulness of lexical phrases in contemporary coursebooks,’ ELT Journal 59/4: 322–32. Kramsch, C. 1988. ‘The cultural discourse of foreign language textbooks’ in A. Singerman (ed.): Toward a New Integration of Language and Culture. Northeast Conference Reports 1988. Northeast Conference. Lakoff, R. 1973. ‘Questionable answers and answerable questions’ in B. B. Kachru, R. B. Lees, Y. Malkiel, A. Pietrangeli, and S. Saporta (eds): Issues in Linguistics. Papers in Honor of Henry and Rene Kahane. University of Illinois Press. Lam, P. W. Y. 2006. ‘Well but that’s the effect of it: The use of well as a discourse particle in talk shows,’ Sprache und Datenverarbeitung (International Journal for Language Data Processing) 30/1: 99–108. Lam, P. W. Y. 2008. Discourse Particles in an Intercultural Corpus of English. Unpublished Ph.D. dissertation. The Hong Kong Polytechnic University. Leech, G. and J. Svartvik. 2002. A Communicative Grammar of English. Longman. Leetch, P. 2002. Use of English Oral Handbook. Free Press. McCarthy, M. 1998. Spoken Language and Applied Linguistics. Cambridge University Press. McCarthy, M., J. McCarten, and H. Sandiford. 2005a. Touchstone. Student’s Book 1. Cambridge University Press. McCarthy, M., J. McCarten, and H. Sandiford. 2005b. Touchstone. Student’s Book 2. Cambridge University Press. McEnery, T. and A. Wilson. 1996. Corpus Linguistics. Edinburgh University Press.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Cheng, W., C. Greaves, and M. Warren. 2005. ‘The creation of prosodically transcribed intercultural corpus: The Hong Kong Corpus of Spoken English (prosodic),’ ICAME Journal 29: 5–26. Cook, G. 1998. ‘The uses of reality: A reply to Ronald Carter,’ ELT Journal 52/1: 57–63. Cook, G. 2001. ‘The philosopher pulled the lower jaw of the hen,’ Applied Linguistics 22/3: 366–87. Davies, A. 2003. The Native Speaker: Myth and Reality. Multilingual Matters. Davies, A. 2004. ‘The native speaker in applied linguistics’ in A. Davies and C. Elder (eds): The Handbook of Applied Linguistics. Blackwell. de Klerk, V. 2005. ‘Procedural meanings of well in a corpus of Xhosa English,’ Journal of Pragmatics 37: 1183–205. Erman, B. 1987. Pragmatic Expressions in English: A Study of You Know, You See and I Mean in Face-to-face Conversation. Almqvist & Wiksell. Esser, D. 1999. Teach & Practice: AS-level Oral English for Form 6. Pilot Publications. Fischer, K. 2006. ‘Towards an understanding of the spectrum of approaches to discourse particles: Introduction to the volume’ in K. Fischer (ed.): Approaches to Discourse Particles. Elsevier. Fraser, B. 1990. ‘An approach to discourse markers,’ Journal of Pragmatics 14: 383–95. Fung, L. and R. Carter. 2007. ‘Discourse markers and spoken English: Native and learner use in pedagogic settings,’ Applied Linguistics 28/3: 410–39. Gilmore, A. 2004. ‘A comparison of textbook and authentic interactions,’ ELT Journal 58/4: 363–74. Graddol, D. 1997. The Future of English? British Council. Graddol, D. 2006. English Next. British Council. Halliday, M. A. K. 1985. An Introduction to Functional Grammar. Edward Arnold. Halliday, M. A. K. and R. Hasan. 1976. Cohesion in English. Longman. Holmes, J. 1984. ‘Hedging your bets and sitting on the fence: Some evidence for hedges as support structures,’ Te Reo 27: 47–62. Holmes, J. 1988. ‘Doubt and certainty in ESL textbooks,’ Applied Linguistics 9/1: 21–44. Hong Kong Examinations and Assessment Authority. 2005. Syllabuses for the Use of English. Advanced Supplementary Level.

P. W. Y. LAM

Romero Trillo, J. 2002. ‘The pragmatic fossilization of discourse markers in non-native speakers of English,’ Journal of Pragmatics 34: 769–84. Schiffrin, D. 1987. Discourse Markers. Cambridge University Press. Schourup, L. 1999. ‘Tutorial overview: Discourse markers,’ Lingua 107: 227–65. Schourup, L. 2001. ‘Rethinking well,’ Journal of Pragmatics 33: 1025–60. Seidlhofer, B. 2002. ‘Basic questions’ in K. Knapp and C. Meierkord (eds): Lingua Franca Communication. Peter Lang. Sinclair, J. and M. Coulthard. 1975. Towards an Analysis of Discourse. Oxford University Press. Sinclair, J. and A. Mauranen. 2006. Linear Unit Grammar: Integrating Speech and Writing. John Benjamins. Stenstro¨m, A.-B. 1990. ‘Lexical items peculiar to spoken discourse’ in J. Svartvik (ed.): The London-Lund Corpus of Spoken English: Description and Research. Lund University Press. Stenstro¨m, A.-B. 1994. An Introduction to Spoken Interaction. Longman. Sutton, M. E. and J. Duncan. 1999. Speaking Precisely 1. Precise Publications. Tribble, C. and W. Jones. 1990. Concordances in the Classroom: A Resource Book for Teachers. Longman. Watts, R. J. 1988. ‘A relevance-theoretic approach to commentary pragmatic markers: The case of actually, really and basically,’ Acta Linguistica Hungarica 38: 235–60. Widdowson, H. 2003. Defining Issues in English Language Teaching. Oxford University Press. Wierzbicka, A. 1991. Cross-cultural Pragmatics. The Semantics of Human Interaction. Mouton de Gruyter. Wong, J. 2002. ‘ ‘‘Applying’’ conversation analysis in applied linguistics: Evaluating dialogue in English as a second language textbooks,’ IRAL 40/1: 37–60. Zorzi, D. 2001. ‘The pedagogic use of spoken corpora: Learning discourse markers in Italian’ in G. Aston (ed.): Learning with Corpora. Athelstan.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Mindt, D. 1997. ‘Corpora and the teaching of English in Germany’ in A. Wichmann, S. Filgelstone, G. Knowles, and A. McEnery (eds): Teaching and Language Corpora. Longman. Mo¨llering, M. 2004. The Acquisition of German Modal Particles. A Corpus-based Approach. Peter Lang. Mu¨ller, S. 2004. ‘Well you know that type of person: Functions of well in the speech of American and German students,’ Journal of Pragmatics 36: 1157–82. Mu¨ller, S. 2005. Discourse Markers in Native and Non-native English Discourse. John Benjamins. Nelson, M. 2000. A Corpus-based Study of the Lexis of Business English and Business English Teaching Materials. Unpublished thesis. University of Manchester. O’Keeffe, A. and F. Farr. 2003. ‘Using language corpora in initial teacher education: Pedagogic issues and practical applications,’ TESOL Quarterly 37/3: 389–418. O’Keeffe, A., M. McCarthy, and R. Carter. 2007. From Corpus to Classroom. Language Use and Language Teaching. Cambridge University Press. Olson, D. R. 1980. ‘On the language and authority of textbooks,’ Journal of Communication 30: 186–96. Pennycook, A. 1998. English and the Discourses of Colonialism. Routledge. Phillipson, R. 1992. Linguistic Imperialism. Oxford University Press. Potter, J. 2003a. Steps & Skills Oral 6. Witman. Potter, J. 2003b. Steps & Skills Oral 7. Witman. Prodromou, L. 2003. ‘In search of the successful user of English,’ Modern English Teacher 12/2: 5–14. Quirk, R. 1955. ‘Colloquial English and communication’ in A. J. Ayer, C. Cherry, B. Ifor Evans, D. B. Fry, J. B. S. Haldene, R. Quirk, G. Vickers, T. B. L. Webster, R. Wittkower, and J. Z. Young (eds): Studies in Communication. Secker and Warburg. Reynolds, P. 1974. English Language Teaching and Textbooks in Hong Kong. Educational Studies and Research Papers No. 3. Department of Education Research Unit, The University of Hong Kong.

281

Applied Linguistics: 31/2: 282–300 ß Oxford University Press 2009 doi:10.1093/applin/amp027 Advance Access published on 29 June 2009

An Extended Positioning Analysis of a Pre-Service Teacher’s Better Life Small Story GARY BARKHUIZEN University of Auckland, New Zealand

Pavlenko (2007) provides an extensive and informative overview of narrative research in applied linguistics, particularly with regard to sociolinguistic analyses of second language users’ personal narratives. The aim of her article is to provide a critical review of the analytical frameworks and procedures used in this body of research, and to offer recommendations for future narrative research work in applied linguistics. Her main concern is that too much analytical emphasis has been placed on the content of narratives, at the expense of their form and contexts of construction. Various approaches to analysis have been proposed in the narrative research literature, some of which directly address the form-content relationship discussed by Pavlenko. Riessman (1993: 2), for example, says that narratives are ‘not merely information storage devices’ containing the content of the narrator’s past life experiences, but in the telling are constructed in particular ways which give meaning to that content. Analysis, therefore, should pay attention to the ‘forms of telling about experience, not simply the content to which language refers’ (Riessman 1993: 2). In presenting his typology of narrative analysis, Mishler (1995: 90) too raises the question of whether the ‘focus of analysis is the actual ‘‘succession of happenings’’ or its textual representation’.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The analysis of narrative data in applied linguistics has focused to varying degrees on their content, form, and context, with content and thematic analyses being the focus in much of the narrative research in language learning and teaching (Pavlenko 2007). The aim of this article is to report on a positioning analysis of a small story about the imagined ‘better life’ of a migrant, pre-service teacher. Positioning analysis operates on three levels, which together require the analyst to examine the content and characters in the story, the interactive performance of the story, and the positions that are agentively taken by the narrator vis-a`-vis normative discourses. Positioning analysis thus considers content, form, and context. I propose and demonstrate an extended version of this approach which enables inclusion of data beyond the small story. The analysis reveals how the teacher interactively constructs an answer to the question ‘Who am I?’ in her story.

G. BARKHUIZEN

283

SMALL STORIES AND POSITIONING ANALYSIS As Watson (2007: 372) says, ‘much of the research on personal narratives focuses on the bigger narratives we tell about ourselves, the stories about our lives that we have reflected on and present as the polished accounts of who we are’. Bamberg (2006: 139) points out that recently, however, there has been an increasing emphasis on narrative practices in narrative research, specifically on the ‘contexts in which narratives take place, what they consist of, their performances, and ‘‘small-story research’’ all seem to be part of this game plan’ (Bamberg 2006: 139). Small stories ‘are the ephemeral narratives emerging in such everyday, mundane contexts, which it is argued constitute the performance of identities and the construction of self’ (Watson 2007: 371). Small stories are small in another sense too; since they are embedded in longer stretches of conversation they are literally smaller or shorter in size. In short, small stories are discursively constructed accounts of identity making. Drawing on Davies and Harre´’s (1990: 48) notion of positioning as a discursive practice ‘whereby selves are located in conversations as observably and intersubjectively coherent participants in jointly produced story lines’, Bamberg (2006: 144–5) defines positioning as follows: In and through talk, speakers establish (i) what the talk is about (aboutness/content), and simultaneously (ii) the particular social

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

In this article I present an approach which draws on and extends the small story, positioning analysis work of Bamberg (1997, 2004a, 2004b, 2004c, 2006), an approach which offers one way of bringing together a focus on content, form and context in the analysis of narrative data. Positioning analysis operates on three sequential levels which together take into account the relationship among characters in the story, the content of the told story, the telling or interactive construction of the story, and the discourses within which narrators position themselves and by which they are positioned in the story world. The purpose of such an analysis is to uncover ‘the process of how . . . positions come into existence and how they assist the construction of a sense of self and identities’ (Bamberg 2004a: 137). Narrative studies of the professional development and practices of second/ foreign language teachers and teacher educators have also, to varying degrees, focused on the form, content and context of narrative data (Casanave and Schecter 1997; Johnson and Golombek 2002; Curtis and Romney 2006; Barkhuizen 2008). A significant branch of this work pays particular attention to teacher’s identity, where the same diversity of analytical approaches can be found (Kramsch and Lam 1999; Tsui 2007). The main narrator of the small story I analyze is a pre-service English teacher who emmigrated to New Zealand from the Pacific Island nation of Tonga before starting her formal teacher education.

284 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

interaction in the form of particular social relationships. And in the business of relating the world that is created by use of verbal means to the here and now of the interactive situation, speakers position themselves vis-a`-vis the world out there and the social world here and now. It is in this attempt of relating aboutness/ content to the social interactants, or making the aboutness/content of their talk relevant to the interaction here and now, through which a position, from where these two ‘worlds’ are drawn together, becomes visible.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Positioning analysis thus attempts to link two approaches (Bamberg 1997: 336): ‘how people attend to one another in interactional settings’ (Davies and Harre´ 1990) and ‘the analysis of what the language is referentially ‘‘about’’, namely sequentially ordered (past) events and their evaluations’ (Labov 1997). To do so, positioning analysis makes ‘the interactive site of storytelling the empirical ground, where identities come into existence and are interactively displayed’ (Bamberg 2004a: 36). The storyteller’s interactive engagement in the story’s construction is thus central to the positioning analysis process. Positioning takes place at three levels, and the analysis focuses on these levels in turn, moving progressively from the local to the global. Level 1 positioning analysis asks the question, ‘How are the characters positioned in relation to one another within the reported events?’ (Bamberg 1997: 337), or as Watson (2007: 374) asks, ‘What is this story about?’ and ‘Who are the characters and why are they positioned this way?’ Put simply, the focus here is very much on the content of the story, who the characters are, and how they relate to each other. Answering these questions typically involves a detailed line-by-line analysis of the linguistic means used to draw up the characters and relate them in time and space. Level 2 positioning analysis asks ‘how the speaker both is positioned by and positions him/herself to the actual or imagined audience. This level concerns how the content and structure of the talk are actually interactive effects’ (Korobov 2001). The focus on this level is on the performance of the story in a particular interactive setting. At level 3, having worked through the interactional context of levels 1 and 2, ‘we are better situated to make assumptions about the ideological positions (or master narratives) within which narrators are positioning a sense of self’ (Bamberg 2006: 145). In other words, ‘whatever has been accomplished locally between the interactants by sharing the story can be told about the speaker elsewhere’ (Bamberg 2004b: 336–7). The analysis thus moves beyond the small story content and telling to consider the normative discourses (the broader ideological context) within which the characters agentively position themselves and by which they are positioned. It seems apt to me at this juncture in the analysis to move outside the small story as data to consider additional narrative data which may be available to the analyst. As Freeman (2006: 135) says, ‘we are not only the selves that issue from small stories. Whether we like it or not, we are also . . . big story selves’. So, instead of ‘bracketing’ these data as suggested in Bamberg’s approach to

G. BARKHUIZEN

285

BACKGROUND Sela was born in Tonga. Early in her life she befriended English-speaking children in her community, and by the time she reached secondary school her level of English proficiency was high enough for her to enroll in a prestigious English-medium school. When she finished school she volunteered to work as a teacher-aide in a rural primary school, where she encountered what she considered to be poor English teaching. She then emigrated to New Zealand and lived in a suburb of Auckland with a high density of Tongan and other Pacific Island immigrants. She soon decided to study to become an English teacher. Sela completed her first degree, followed by a postgraduate diploma in language teaching and learning, and then an MA in the same subject. It is important to note for my later discussion that both these graduate qualifications had a curriculum emphasis on language teaching and learning—not only or specifically on English. I met Sela during the early stages of her postgraduate studies. She enrolled in two of the courses I taught, and then when she reached the research dissertation stage of her MA, I became her supervisor. Sela is now in her mid-20s, definitely wants to be an English teacher within her own community in Auckland, and her English proficiency is high. The idea for this study emerged during our discussions on the topic of Sela’s dissertation. She conducted talanoa1 interviews with three Tongan women to investigate their post-immigration English learning experiences in New Zealand. Sela was particularly interested in exploring in her data the concepts of investment and imagined communities (Norton 2001; Pavlenko 2003). Many of the issues raised in their stories were similar to those Sela experienced in her own life. In one of the courses she had studied with me, Sela was

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

positioning analysis, and as one would expect in conversation analysis (Schegloff 1997; Korobov 2001), I propose extending the analysis only at level 3 to include other data which may have been collected during the course of any narrative inquiry project. This may include other small storiesin-interaction collected from the same participants at different times in different settings, a move which, Bamberg (2004c: 369) concedes, ‘can lay out revisions in our sense-making and re-storying capabilities in much more detail and with greater effectiveness’. Additional data could also include ethnographic data collected in the form of interviews and written journals. Some analysts may find it very difficult to ignore such data, especially when relating them to broader discourses, and would possibly argue that doing so till level 3 is restrictive enough. Extending the analysis at level 3 is not a compromise, however. Instead, I suggest that it is the most opportune stage in the analysis to do so, and without foregoing the detailed (bracketed) textual analysis at levels 1 and 2. In the next section, I provide background information regarding the small story and the main participant in the study as well as details about the big story data and how it was collected.

286 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

POSITIONING ANALYSIS LEVEL 1 Level 1 analysis of the small story pays ‘close attention to the ways in which the constructed/represented world of characters and event sequences is drawn up’ (Bamberg 2006: 145). I started the analysis by generating a detailed list of all the characters (real and imagined) in the story. Doing so, through a careful line-by-line reading, enabled me to picture and to begin to know the people Sela relates to in the story, including those imagined in her future teaching life. Re-reading the story a number of times together with the list of characters (and making margin notes along the way) was important for understanding more deeply the nature of the relationships Sela constructed with the other characters in the story. I then turned my focus to the events in the story. By events I mean the content of the story; i.e. what happens in the story, or as Bruner (2006: 133) puts it, the ‘landscape of action on which events unfold’, and I include in the list Sela’s evaluations (Labov 1997, 2006) of these events. The purpose of generating the events list is to ensure that, firstly, a deeper awareness and understanding of the events is reached, and secondly, that nothing is missed. Sela is the main character in the story and is thus deeply involved in the action, both in terms of what has happened and what she imagines happening in the future. As narrator she focalizes the story world of the characters and events (Toolan 2001), providing a description and evaluation of these events and how the characters are ordered in space and time inside the world of the story, and by so doing positions the characters in the story, including herself, in relation to that story world. Once again, a careful, repeated reading of the story together with the list of events (and making margin notes along the way) was important for understanding the story and

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

required to write a series of four personal reflective narratives, which she then analyzed for the purposes of a class assignment. I decided, with Sela’s consent, to examine these further, and also to conduct a series of three narrative interviews (Atkinson 2002; Chase 2003) with her over a period of 8 months. The aim of the study was to investigate Sela’s emerging teacher identity as a preservice English teacher imagining her future working life. In addition to the four written narratives and the three interviews, Sela and I met regularly, during the same eight month period, to discuss progress on her dissertation. These conversations included sharing many stories of our ideas about teaching, about living as migrants in New Zealand, and about language learning. The above formed the big story data set for the study. The project was approved by the ethics committee of my institution. The small story (see appendix) is an extract from our second interview. In the story Sela tells of two related conflicts which she believes are central in preventing access to effective English learning for Tongan adult migrants living in Auckland. Her own (imagined) part in overcoming these conflicts when she starts to teach is also revealed in the story.

G. BARKHUIZEN

287

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

answering the question ‘What is this story about?’ (Watson 2007: 374). Answers to this question clearly focus the analysis on the content of the story, but the content is ‘reached’ through a detailed analysis of the structure of the story and its conversational units (Korobov 2001). My approach, as described above, may not utilize the precise sort of conversation analytical tools that Bamberg (1997, 2004b) suggests. It does, however, at this level, focus the analysis on both the content and form of the story, as suggested by Pavlenko (2007),2 and also makes an empirically grounded start to the analysis (Bamberg 2004a). What, then, is this story about? Who are the people in the story and how do they relate to one another? In what follows, I make Sela, the focalizing narrator, the locus of the discussion, since the purpose of the analysis is to uncover the claims she discursively makes about her developing teacher identity. Right at the start of the small story, Sela positions herself both as a pre-service teacher and as an affiliate of the Pasifika, particularly the Tongan, community. She talks about the ‘people on my street’ and her ‘aunties or my cousins’ who will attend her English classes. Throughout the story she makes reference to ‘Tongan adults’ (line 14), ‘Tongan adult immigrants’ (line 16), and ‘a class specifically for Tongan and Pasifika’ (line 72). These are the people amongst whom she lives her life, and the people she ‘ideally’ (line 15) desires to teach. In the story Sela displays a keen awareness of the cultural values and practices of this community. Her imagining ‘a little like night class in our garage’ (line 9) and her ‘threat’ to ‘force my family to come’ to her English classes (line 79) seems to demonstrate her knowledge of the ‘collective processes in the way that Pacific peoples operate’ (Vaioleti 2006: 32). She knows what are culturally inappropriate classroom practices for Tongan learners: see lines 35–55, for example, where she tells a story of an ‘older’ (line 42) Tongan woman and her cousin who encountered the word ‘shit’ in a text they were working with in class, and consequently felt ‘quite embarrassed’ (line 43) and ‘quite uncomfortable’ (line 55). She is aware of inappropriate inter-gender behavior (lines 28–34), and she is able to explain and comment on the (from her perspective) destructive consequences of the ma barrier; an obstacle in the way of engaging with English language learning. Sela defines the ma phenomenon as ‘shame’ (line 85) and ‘quite an embarrassing sort of thing’ (line 90), which surfaces when the potential exists for a person to be identified as not being able to speak English, and thus considered ‘quite dumb’ (line 89) and ‘uneducated’ (line 91). Sela’s constructed positions as both pre-service teacher and staunch member of the Tongan community make possible one of the two major conflicts (conflict 1) in the story. This conflict, which pervades the entire small story, concerns the culturally inappropriate English curriculum experienced by Tongan adult learners. Sela is aware of this conflict and she explicitly states what she wants to do about it: ‘to do things like constructing curriculum or syllabus specifically for Tongan adult immigrants’ (line 16). In this way

288 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

she positions herself as mediator between the opposing forces that constitute the conflict; which are, on the one hand, the English curriculum, which she will work with and help to develop in her imagined teacher life, and, on the other, her position as a culturally aware member of the Tongan community, from which will come her imagined English learners. Conflict 1 represents the main driver in Sela’s imagining being a teacher. Working towards resolving the conflict is what gives her future work a purpose and impacts the ongoing construction of her imagined teacher identity. Conflict 1 is directly related to the second major conflict in the story (conflict 2). This conflict concerns the ma barrier which Sela believes prevents Tongan adults from enrolling in English language classes. There would be no point in Sela trying to resolve conflict 1 if she taught in circumstances where her students were not those she desired to teach: i.e. Tongan adult immigrants. In other words, the relevant learners need first to go to English classes. Sela’s imagined work in getting them there positions her yet again as a mediator between the constituents of conflict 2, namely, going to class to learn English and the ma phenomenon. Her position as mediator in conflict 2, however, has a somewhat different quality than her mediator position in conflict 1. This time a sense of activism is evident in her desire to deal with conflict 2. Her cause is enabling potential learners to overcome the ma barrier, and she imagines a number of ways of doing so. She will ‘so force her family to come’ to English classes (line 80), for example, and will do whatever she can to ‘encourage’ (line 94) members of her community to learn English because it will ‘improve your life’ (line 97) and ‘will be good for you living here in New Zealand’ (line 100).3 In Sela’s depiction of conflict 2, she uses present tense to position her imagined learner characters as disadvantaged migrants living in New Zealand in the here and now; e.g. ‘if your English is not good enough’ (line 88), ‘so you’re not just dumb, you’re uneducated’ (line 91), and ‘it’s a barrier to Tongans’ (line 102). She switches tense to make future references when exposing the harmful consequences of ma, positioning the Tongans adults as unfortunate recipients of its power; e.g. ‘people will say things’ (line 106), ‘so-and-so down the road will actually realize how dumb or uneducated they feel she is’ (line 111), and ‘someone won’t go to English class’ (line 110). And Sela continues with future references to express the desired outcomes of her imagined teacher work—e.g. ‘if I could encourage them that it’s not about that’ (lines 94–5), and ‘this will be good for you living here in New Zealand’ (line 100). In the story Sela makes connections, therefore, between the undesirable current circumstances of disempowered Tongan migrants in her community, and, through her imagined mediation work as an activist teacher, these same people enrolling in English classes, learning English and thus improving their level of education and their life in New Zealand. In sum, this story is a ‘better life’ story about the people in Sela’s community, and the part she plays in contributing to bring about this better life.

G. BARKHUIZEN

289

POSITIONING ANALYSIS LEVEL 2

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Positioning analysis level 2 focuses on the performance aspect of the story; ‘the interactional means employed for getting the story accomplished’ (Bamberg 2004a: 137). Since the small story is an extract from a narrative interview, it is constructed by both Gary4 the interviewer, and Sela. Because of the nature of narrative interviews, where participants are invited to ‘tell me about’ (Chase 2003: 88), and during which stories are then told, it is to be expected that Sela holds the floor for much of the extract. Gary’s role in the interview could best be described as supporter (i.e. keeping the conversation going) and prompter (i.e. digging deeper), typical of narrative interviews with a research agenda. Gary begins (lines 1–7) by asking Sela to address the topic raised immediately prior to the opening of this extract; i.e. Sela’s plans for where and in what sort of teaching arrangement she desires to teach. The informal tone of his complex question (e.g. line 6, ‘you can’t just knock on people’s doors’) is illustrative of the relaxed, open rapport built up between them over the duration of their narrative work together. Her response, which is the start of the small story, begins in line 8. The ‘um’ followed by ‘the one thing that keeps coming up in my mind is’ signals a story opening. The opening scene is an imagined night class in Sela’s garage, something she had ‘originally’ considered, indicating that her thoughts about this have since changed; to what, exactly, remains unclear, to both her and Gary, and is not articulated explicitly throughout the rest of the small story. Gary’s question which launched the story is therefore never actually answered. Sela holds the floor for almost all of lines 8–25, during which she introduces her imagined English learners; what Toolan (2001: 62) refers to as ‘imagined focalizeds’. These include those who would attend her night classes (‘people on my street’, ‘whoever’, her ‘aunties or my cousins’), and more likely, Tongan adult immigrants. Also in lines 8–25 Sela introduces conflict 1; the culturally inappropriate curriculum experienced by Tongan adult learners. She begins by stating one of her imagined future teaching goals, which is to construct ‘curriculum or syllabus specifically for Tongan adult immigrants’ (line 16), but is not specific about how she plans to actually do this. She refers to conversations she’s had ‘with people who’ve taken any English courses’ which are the source of information on which she bases her knowledge of the conflict. Line 25, ‘you know’, suggests to Gary that Sela has finished addressing conflict 1 and, being unsatisfied with his understanding thereof, asks (line 26) for an example in order to interpret the conflict more fully. Sela obviously had one in mind since she immediately launches into a story about such a case (line 27), but interrupts it after only one line to insert a sociocultural context for the story (lines 28–34). The description of this context is brief and not very coherent, but its aim is to draw Gary’s attention to Sela’s understanding of some relevant (particularly inter-gender) constraints within Tongan cultural practices,

290 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

signaled by negative phrases such as ‘can’t do anything’, ‘don’t say anything bad’, ‘don’t watch’, and ‘a big taboo’. Line 35 continues the example, providing the sort of in-depth story content that Gary requested in order to exemplify conflict 1. The example, which ends at line 55, implicitly shows Sela positioned as an empathizer, with the key ‘lady’ learner in the story, the lady’s cousin, and generally the ‘older’ Tongan English learners in the class being the target of her empathy. Her feelings of empathy, however, no doubt also extend beyond the actual class members in the example to the imagined Tongan adult immigrant learners Sela plans to teach in the future. In this way, the example informs Gary of Sela’s equity and pedagogical concerns, which are the basis of conflict 1. In line 56 Gary asks if there is ‘another way of doing it’, referring to a teaching arrangement only for Tongan learners which would go some way towards solving the conflict. Sela’s answer (lines 59–68) reveals that she is unaware of any such existing arrangement, but that she is aware of schools in her community ‘out south’ which offer Tongan language classes for Tongans. Her answer is therefore a comment on Tongan language maintenance practices, ending with the ironic statement in line 68; ‘it’s like all Tongans know how to speak Tongan’. The reference to Tongan maintenance may seem like an aside to this story, but it provides details of the complex sociolinguistic (migrant) context within which Sela imagines herself teaching, and perhaps also, while telling the story, allows Sela more time to consider an apparently more relevant answer to Gary’s question in line 56. This answer (lines 70–7) is also not satisfactory, so Gary asks Sela (line 78) to imagine with more specificity the people whom she will teach. The result is remarkable. In her response Sela takes up the position of an enforcer, or what I prefer to call an activist. Without hesitation she indicates that she would force her family to attend the classes, and then stresses the point by repeating the statement, this time adding the intensifier ‘so’ (line 80). She then tells Gary that she believes more Tongans actually want to learn English than actually do, but their desire is mitigated by the cultural phenomenon called ma, discussed earlier in the level 1 analysis. Ma operates as a barrier to learning, which, she says, is ‘revered by Tongans’. And knowing English represents education. The second conflict is thus introduced into the story. In lines 88, Sela suddenly begins using second person pronouns; ‘your English’, ‘how smart you are’, and continues to do so through to line 91. The tone of this turn is very much like that of a sermon or lecture in which she appears to be directly addressing those who succumb to the ma barrier. There is a short break in the sermon (lines 92–6) where Sela again speaks to Gary, re-stating the meaning of ma and the related conflict 2, and positioning herself, in her imagined teacher identity, as mediator of this conflict; ‘so if I could encourage them that it’s not about that’ (lines 94–5). The sermon continues in line 97 (through to line 106), and in this same line articulates succinctly what sort of small story this is; i.e. a better life story. Sela’s moral position that she displays to Gary (and the non-present Tongan addressees of her sermon) is that actively

G. BARKHUIZEN

291

engaging in English learning (which equates to ‘how educated you are’) and thus overcoming the ma barrier will lead to a better life. The use of second person pronouns continues throughout the sermon climaxing with the accusatory ‘and you won’t do it because you may fail’ (line 105). Gary repeats this message in line 107, also using second person, and so does Sela in line 108, together concluding the sermon.

POSITIONING ANALYSIS LEVEL 3: WHO AM I?

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

In level 3 analysis I examine how Sela transcends the story content and the interactive story telling to address the question ‘Who am I?’ and ‘Who do I imagine myself to be in the future?’ in relation to broader discourses in the world out there. In extending this level of analysis, I argue that it is legitimate to consider additional data—data from Sela’s big story—in order to reach a more comprehensive understanding of Sela’s identity positions (constructed and displayed in the small story, and now also evident in her big story). Sela’s narrative orientation brings into focus the central theme of a ‘better life’, for herself as well as the other characters in the story. The analysis so far has shown how Sela agentively positions herself within a pervasive discourse of immigration, but she is also positioned by this discourse. Who she is, is rooted in a discourse which positions migrants as disadvantaged in a number of respects, including linguistically, economically, educationally, emotionally and professionally (de Fina 2003; Yelenevskaya and Fialkova 2003; Barkhuizen and de Klerk 2006). Sela’s main point in the small story is that through learning English her learners will have the opportunity to gain access to some of the capital that New Zealand society has to offer. Investing in learning English (Norton 2001), therefore, will lead to a better life. At the same time, Sela’s investment in teacher education and later in her teaching means that she too can draw on that capital in the future. Through studying to be a language teacher Sela engages with a discourse of language teacher education, currently being positioned as a pre-service teacher imagining her work in the years to come. Her desire for this investment is to yield an improved quality of life for herself and her family. Positioned, then, within the two dominant discourses of immigration and language teacher education, Sela makes identity claims about herself, who she is and who she wants to be. Underlying both discourses is an ELT ideology (Pennycook 1998; Canagarajah 1999) which has obviously inspired Sela: English is the language of power in New Zealand, and gaining access to what New Zealand society has to offer (e.g. employment, financial security, a good education) necessarily requires proficiency in English. Progressing to a better life, in Sela’s terms (for herself and her English learners), has sociopolitical implications, however. For example, what about those who do not desire to take up English learning (those Sela would ‘force’ to enroll in English classes)? Why don’t they? And, what are the sociopolitical costs for those who do? I address these questions within the context of four positions which the analysis generated.

292 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

Tongan migrant

Teacher Sela has no teaching experience, but already is thinking about her future life as a teacher. She positions herself within an imagined community (Norton 2001; Pavlenko 2003) of teachers, teaching institutions, and English learners. In Sela’s small story she connects with a number of imagined English learners

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Sela is a Tongan migrant; she lives in a Tongan community in New Zealand, knows the community and its cultural practices, and wants to contribute to it by practicing as a teacher within it. This is all evident in the small story. What her big story adds is her history; her story of growing up in Tonga, going to school, and learning English. In her written narratives, for example, she portrays rich descriptions of the context in which she spent the early part of her life. These include a picture of the socioeconomic divide within the Tongan school system, reflecting the structure of Tongan society more generally. Sela learned her English in the playground (what she calls ‘playground learning’) mixing with ‘the kids of academics and professionals who went to the prestigious English speaking school’. She did not attend their primary school but by the time she reached secondary school her English was good enough for her to be accepted into ‘the best school in Tonga’, which was attended by all her friends and children from the expatriate community. Sela realizes she is privileged because of her high level of English proficiency, saying ‘I was the lucky one’, adding that children in rural school classrooms have little chance of passing English—a requirement to ‘progress to the next level’—and thus of finishing school. Her awareness of the same dilemma is similarly acute in the New Zealand immigrant context, as the small story analysis has shown. In one of her written narratives, for example, Sela contemplates the unfortunate position of children in her community who ‘learn bad Tongan, bad English and mixed up ungrammatical English’ from their parents and from rap music and hip hop, and gives this as a reason why ‘many kids in south Auckland drop out of school . . . there is no way that they are going to get very far in life, to get a decent job, at least with that sort of English’. For Sela, the route out of this position of disadvantage is primarily economic, based on improved English proficiency and a better education. But there may also be sociopolitical consequences to these outcomes which are not explicitly told in her story (big and small): Sela is positioned within and by an immigration discourse, and this is a discourse which includes not only the prospect of social and economic redemption through English, but also resistance to the power of the host community, adaptation of what it has to offer to suit themselves, and rejection of many of its practices (Pennycook 1998; Ibrahim 1999). As a future English teacher working within such an immigrant community Sela will, as she is aware, certainly encounter these consequences.

G. BARKHUIZEN

293

Activist/mediator What drives Sela’s desire to be a teacher is her ‘deep underlying motivation to help my family and my community’. In order to meet this goal Sela takes up an activist teacher position, in her case mediating between the elements constructing the two conflicts in the small story (as described above). At the heart of these conflicts are the inequitable positions of the learner immigrants in the story world with regard to the access they have to symbolic and material capital (Norton 2001). In Sela’s story, she constructs the migrant characters as disadvantaged within an immigration discourse. Her big story is peppered with statements which show this orientation. For example, in the last interview Sela describes the experiences of three recent Tongan migrant women

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

who will play a vital part in her teaching life. Her big story data confirm these to be her desired learners, and also reveal who she does not want to teach (‘not high school’). She imagines where she will teach: e.g. ‘a little like night class in our garage’ (line 9), and ‘a class specifically of Tongan and Pasifika’ (line 72). Outside the small story Sela imagines teaching abroad, saying ‘I thought maybe I could go and take five years out and go travel and I’m seeing this investment [her teacher education] as my ticket to do it’.5 Ultimately, though, she sees herself teaching within her south Auckland community, to ‘live here and teach here forever, for a very long time’. Pavlenko (2003: 253) points out that the identitary function of imagination ‘allows us to view appropriation of newly imagined identities as an important aspect of a learning trajectory, which transforms apprentices or peripheral community members into legitimate participants’. In other words, Sela, through claiming for herself an imagined teacher identity in her narrative (big and small) and within a language teacher education discourse, is undergoing teacher development which is impacted upon by her imagining a teaching community, including the immigrant community in which she lives and will work. Sela is also positioned as teacher in her narrative tellings of her past experiences. In her big story Sela recalls in an interview the two years she worked as a volunteer teacher-aide in a rural Tongan school after graduating from high school. Her observations of the teaching ‘which took place class-afterclass-after-class’ led her to declare, ‘I do not, I do not want to teach like that’. She describes it as follows: ‘The main form of teaching, the teacher would write the notes on the blackboard and then explain. When you have finished writing off all the notes, then they would explain. . . . They’ll explain it in Tongan. It was just crazy’. Sela says she is ‘not surprised a lot of those kids failed English’, and is determined not to teach in the same way. Sela clearly positions herself, then, within a language teacher education discourse. Her developing teacher self simultaneously operates within an immigration discourse, as has been shown, and so the constant references to teaching ‘whanau’ [extended family], ‘my cousins and brothers’, Tongans, and ‘people on my street’ are to be expected.

294 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

Investor/capitalizer The identity position ‘investor’ is based on Norton’s (2000) concept of investment. With reference to language learning, she argues that when learners invest in an additional language they ‘expect or hope to have a good return on that investment—a return that will give them access to hitherto unattainable resources’ (2000: 10). Sela has invested in language teacher education, with the expectation that her university studies and specifically her teacher training will yield returns for herself, her family and the Tongan immigrant community. In getting as far as she has, Sela has made good use of the resources and opportunities available to her. These include learning English to a high level of proficiency by mixing with the (right) friends she grew up with in Tonga and by attending an English-only school, and then by studying English at university in New Zealand. Sela obtained her first degree, followed by language teaching qualifications. She has therefore accumulated a substantial amount of capital relevant to her imagined teacher identity. As capitalizer,7 a position closely associated with investor, and as the three-level positioning analysis has shown, Sela plans to make good use of the knowledge and skills she has acquired when she starts her teacher work, albeit in the face of potential sociopolitical conflicts in her community.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

living in Auckland.6 The following statements are indicative: ‘she was just not satisfied with the job she could get here’, ‘financially they were struggling a lot’, and ‘they also wanted to be able to just fit in’. These circumstances Sela believes are unfair, and she wants to do something about it: ‘It’s a big thing for me, as a teacher’. Sela’s activism has roots in her earlier life experiences growing up in Tonga. She was aware of the sociopolitical structure of Tongan society, and how this was reflected in the school system and the related opportunities students had for learning English. She was also aware of the power that knowing English gave to those proficient in the language. In one of her written stories she tells of her experiences of going to the same high school as two of the Tongan king’s granddaughters. She says that when she (and other student ‘commoners’) spoke to the granddaughters in English she could use ‘normal’ English, but if they spoke Tongan ‘the whole class structure of Tongan society would come into full play’, and the relevant dialect would have to be used. She elaborates in one of her written narratives: ‘I guess the English language for me is very powerful. It is the only time I can think of myself as (almost) equal to nobility’. The social class structure in the New Zealand immigrant context is, of course, different from that in Tonga. Nevertheless, within the Tongan immigrant community social practices reminiscent of those in the home country do possibly still exist, especially among older generations. Perhaps this is why, in addition to or instead of the ma barrier, some Tongans adults are reluctant to attend English classes. This is a dilemma Sela’s activism will no doubt confront.

G. BARKHUIZEN

295

CONCLUSION

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

This article has presented a report on an analysis of a small story, which has drawn on and extended the typical procedures of positioning analysis. I use the words ‘drawn on’ purposefully here, since my approach has not strictly adhered to the analytical procedures prescribed by Bamberg (1997). It has differed in two main ways. First, analyzing the linguistic means used to establish the characters and events in the story, especially at positioning level 1, has probably not been as fine-grained as positioning analysis, which itself draws on the procedures of conversation analysis, would demand. However, the detailed line-by-line examination of the story when searching for characters and events (and generating the lists of these) certainly required me to pay close attention to the text; i.e. how Sela narrated her spatio-temporal orientation of the story world. And the level 2 analysis clearly demonstrates how linguistic resources were used to construct interactively both the story and the local understandings relevant for both participants. The analysis, therefore, answers Pavlenko’s (2007) call for analytic attention to be paid to both the content and form of narrative texts. At levels 1 and 2 context ‘is treated both as the project and product of the participants’ own actions and therefore as locally produced and transformed at any moment’ (Korobov 2001). But at level 3, broader normative discourses come into play in the analysis of identity claim-making, taking into account the sociopolitical context of which the small story is a part, an approach more typical of critical discourse analysis. It is at this level that I propose extending the range of data for analysis, if available. As Freeman (2006: 131) says, ‘big stories and small stories . . . complement one another; taken together, they represent a promising direction for narrative inquiry’. This is the second main way in which I diverge from standard positioning analysis procedure. Doing so means that context is considered at the macro or global level (Pavlenko 2007), not only in terms of broader discourses (typical of level 3 positioning analysis), but also in terms of the historical, political and cultural circumstances of the narrator’s story evident in the big narrative data. Considering these data can only enhance the effectiveness of the analysis, as I hope I have demonstrated above. Working through the three levels, then, demands a focus on content, form, and context, and thus overcomes the limitations of content and thematic analyses described by Pavlenko (2007). Most important, of course, is achieving the aims of the analysis; i.e. to investigate identities emergent in small stories-in-interaction, and to answer the question ‘Who am I?’ Sela read my analysis as presented above, and she responded with two revealing comments. Firstly, she said ‘I’ve just finished reading through the analysis, and my word, thank you for dissecting my identity complexity(ies) and thus giving me the ability not only to understand it, but also verbalize it. It was VERY INTERESTING reading indeed!’; adding that she thought the analysis was ‘really good’. Her response is satisfying in that it confirmed my analysis, but what is particularly interesting is Sela’s

296 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

ACKNOWLEDGEMENTS I am grateful to Penny Hacker and Zane Goebel for their feedback on a draft of this article, and to the anonymous reviewers for their very useful comments.

APPENDIX: THE SMALL STORY 1 G 2 3 4 5 6 7 8 S

. . . Have you thought about how you’re actually going to go about teaching in terms of where you teach and how you teach? you know you talk about your community and so on but now where? you can’t just knock on people’s doors I mean what sort of institution do you want to get involved in? um the one thing that keeps coming up in my mind is

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

comment about ‘verbalizing’ the analysis for herself. In other words, through making sense of her own claims about identity in the small story, as articulated in my analysis, she is now able to re-story her experiences of becoming a language teacher; and with each re-storying continue repositioning and reimagining her own teacher identity. The implications for teacher education here are obvious. Agee (2006: 212) suggests, for example, that providing spaces for teachers to ‘examine their imagined roles within a community of peers and to hold those roles up to the mirror of current theory and practice’ allows them to evaluate and rethink their ideas about teaching and their imagined roles. Secondly, Sela also responded to the analysis by saying ‘however, I do feel kinda, I don’t know the word, sort of like a pessimist that I make ‘all’ Tongan migrants sound like they’re non- or limited-speakers of English, and hence disadvantaged; after all, there are a lot of Tongan migrants who are highly qualified and proficient in English’. This comment serves as a reminder to narrative analysts that narrators are selective in what they tell (Clandinin and Connelly 1998), and that any analysis, of both small and big stories, should ask questions about what is not told. In Sela’s case, for example, we do not hear explicitly in her story about Tongan immigrants who resist learning English, or about those who are highly qualified and successful in their lives but do not have a high level of English proficiency, or about those who enroll in English classes and are not subsequently successful, in Sela’s terms. Sela’s story appears to reveal how she has positioned herself as complicit with a dominant ELT ideology, expressed particularly as an economic metaphor of investment, capitalization and a better life. But what she does not tell, and which she will surely confront in her future life as an English teacher, are counter-ideologies that will present her with challenges both in her work and when making claims about who she is.

G. BARKHUIZEN

9 10 11 12 13 14 15 16 17

22 G 23 24 S 25 26 G 27 S 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

I had originally thought about having a little like night class in our garage with whoever people on my street or my aunties or my cousins who want to improve their English that’s what I really want to do but if I wanted to teach in terms of like an institution I would most likely like to teach Tongan adults I would ideally want to teach want to do things like constructing curriculum or syllabus specifically for Tongan adult immigrants and that’s just coming from conversations I’ve had with people who’ve taken any English courses and I mean the activities and everything is interesting and they’ve really enjoyed the classes but sometimes the content and materials used they did not quite understand and they’re not quite in sync with their purpose of why they were learning English that will be interesting to explore in your interviews for those who have done those classes and also it’s like there’s some content that’s quite culturally inappropriate for Tongans as such you know what are some examples of that? it’s like they were given a text because like the Tongan social structure there was like we can’t do anything and stuff and you know don’t say anything bad and don’t watch TV together it’s a big taboo on you hanging out with boys with your boy cousins and I think this lady had a cousin distant but a cousin within the same class it was just this text that had sort of like slang and not so much swear words as such but like shit and stuff and she was really because they were older she was like quite embarrassed so it’s yeah just little things like that I mean I think that if it was a younger student of a younger age it wouldn’t have bothered them as much but just because they were much older she was just saying some of the content in the class I don’t know what it was I think the lesson was just to familiarize them with some slang colloquial language

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

18 19 20 21

297

298 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

52 53 54 55 56 G 57 58 59 S 60

G S

G S

71 72 73 74 75 76 77 78 G 79 S 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

61 62 63 64 65 66 67 68 69 70

something like that and I think shit came up once yeah but even then even then for her was quite uncomfortable so is there another way of doing it? I mean what about is there some sort of community centre that could be used specifically for Tongans or? I’m not sure I know that in some of the local schools out south they run language classes but it’s more Tongan language a huge push for maintenance I think there’s more Tongan language classes than for Tongans? for Tongans growing up here {laughs in voice} that sounds funny Tongan language classes for Tongans it’s like all Tongans know how to speak Tongan no, of course not but I know like at [name of a tertiary institution] they offer English language classes but it’s not specifically for Tongans but I wouldn’t mind teaching in a class specifically for Tongan or Pasifika I don’t know I just think that would make them more comfortable to come in and learn English that’s what I think I’m not sure who do you think would come to those classes? I’d force my family to come I would so force them to come I think more Tongans actually want to learn English it’s just that they have this Tongans have this ma called ma or just shame because English to Tongans qualifies as education and education is so highly revered by Tongans that if your English is not good it doesn’t matter how smart you are in any other aspect if your English is not good then you’re considered quite dumb and that’s quite an embarrassing sort of thing so you’re not just dumb you’re uneducated maybe a lot of them wouldn’t go to these things because they’re ma they don’t want to show other people how uneducated they are so if I could encourage them that it’s not about that you know

G. BARKHUIZEN

112 113 114

it’s just to help you improve your life and it’s not about how educated you are or what grade you got up to in high school this will be good for you living here in New Zealand yeah because that ma barrier it’s a barrier to Tongans improving their life there’re so many things they want to do like so many things you want to do and you won’t do it because you may fail and people will say things so you don’t do it you don’t do it that’s the point you know like someone won’t go to an English class because so-and-so down the road will actually realize how dumb or how uneducated they feel she is and so they won’t go it’s a barrier they need to get over in order to be able to learn English or actually go to English language class

NOTES 1

2

3

Tala means to inform, tell, relate, and ask, and noa means of any kind, ordinary and nothing in particular. Talanoa refers to ‘a personal encounter where people story their issues, their realities and aspirations’ (Vaioleti 2006: 21). It does, of course, also focus on context, but local context produced at any interactional moment, as in conversation analysis (Schegloff 1997). The way in which she articulates this activist streak to Gary in the story will be discussed in level 2 analysis.

4

5

6 7

In this section I refer to myself in the third person in order to distinguish between Gary the co-narrator of the story and Gary the author of this article. Changed family circumstances later meant that her ‘fantasy’ to go abroad and travel and teach English did not and will not materialize. These are women she interviewed for her MA dissertation research. I am indebted to Alice Chik for suggesting this term.

REFERENCES Agee, J. 2006. ‘What kind of teacher will I be? Creating spaces for beginning teachers’ imagined roles,’ English Education 38/3: 194–219. Atkinson, R. 2002. ‘The life story interview’ in J. F. Gubrium and J. A. Holstein (eds): Handbook of Interview Research. Sage, pp. 121–40.

Bamberg, M. 1997. ‘Positioning between structure and performance,’ Journal of Narrative and Life History 7/1–4: 335–42. Bamberg, M. 2004a. ‘Positioning with Davie Hogan: Stories, tellings, and identities’ in C. Daiute and C. Lightfoot (eds): Narrative Analysis: Studying the Development of Individuals in Society. Sage, pp. 135–57.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

97 98 99 100 101 102 103 104 105 106 107 G 108 S 109 110 111

299

300 A PRE-SERVICE TEACHER’S BETTER LIFE SMALL STORY

Forum: Qualitative Social Research [On-line journal] 2/3, available at http://www.qualitativeresearch.net/fqs-texte/3-01/3-01korobov-e .htm. Accessed 15 March 2008. Kramsch, C. and W. E. Lam. 1999. ‘Textual identities: The importance of being nonnative’ in G. Braine (ed.): Non-native Educators in English Language Teaching. Lawrence Erlbaum, pp. 57–72. Labov, W. 1997. ‘Some further steps in narrative analysis,’ Journal of Narrative and Life History 7: 395–415. Labov, W. 2006. ‘Narrative pre-construction,’ Narrative Inquiry 16/1: 37–45. Mishler, E. G. 1995. ‘Models of narrative analysis: A typology,’ Journal of Narrative and Life History 5/2: 87–123. Norton, B. 2000. Identity and Language Learning: Gender, Ethnicity and Educational Change. Longman. Norton, B. 2001. ‘Non-participation, imagined communities and the language classroom’ in M. Breen (ed.): Learner Contributions to Language Learning: New Directions in Research. Pearson Education, pp. 159–71. Pavlenko, A. 2003. ‘ ‘‘I never knew I was bilingual’’: Reimagining teacher identities in TESOL,’ Journal of Language, Identity, and Education 2/4: 251–68. Pavlenko, A. 2007. ‘Autobiographic narratives as data in applied linguistics,’ Applied Linguistics 28/2: 163–88. Pennycook, A. 1998. English and the Discourses of Colonialism. Routledge. Riessman, C. 1993. Narrative Analysis. Sage. Schegloff, E. A. 1997. ‘Whose text? Whose context?’ Discourse and Society 8: 165–87. Toolan, M. 2001. Narrative: A Critical Linguistic Introduction. 2nd edn. Routledge. Tsui, A. 2007. ‘Complexities of identity formation: A narrative inquiry of an EFL teacher,’ TESOL Quarterly 41/4: 657–80. Vaioleti, T. M. 2006. ‘Talanoa research methodology: A developing position on Pacific research,’ Waikato Journal of Education 12: 21–34. Watson, C. 2007. ‘Small stories, positioning analysis, and the doing of professional identities in learning to teach,’ Narrative Inquiry 17/2: 371–89. Yelenevskaya, M. N. and L. Fialkova. 2003. ‘From ‘‘muteness’’ to eloquence: Immigrants’ narratives about languages,’ Language Awareness 12: 30–48.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Bamberg, M. 2004b. ‘Form and functions of ‘‘slut bashing’’ in male identity constructions in 15-year-olds,’ Human Development 47: 331–53. Bamberg, M. 2004c. ‘Talk, small stories, and adolescent identities,’ Human Development 47: 366–9. Bamberg, M. 2006. ‘Stories: Big or small: Why do we care?’ Narrative Inquiry 16/1: 139–47. Barkhuizen, G. 2008. ‘A narrative approach to exploring context in language teaching,’ English Language Teaching Journal 62/3: 231–9. Barkhuizen, G. and V. de Klerk. 2006. ‘Imagined identities: Pre-immigrants’ narratives on language and identity,’ International Journal of Bilingualism 10/3: 277–99. Bruner, J. S. 2006. In Search of Pedagogy, Volume 2: The Selected Works of Jerome S. Bruner. Routledge. Canagarajah, S. 1999. Resisting Linguistic Imperialism in English Teaching. Oxford University Press. Casanave, C. P. and S. R. Schecter. 1997. On Becoming a Language Educator: Personal Essays on Professional Development. Lawrence Erlbaum. Chase, S. E. 2003. ‘Learning to listen: Narrative principles in a qualitative research methods course’ in R. Josselson, A. Lienlich and D. P. McAdams (eds): Up Close and Personal: The Teaching and Learning of Narrative Research. American Psychological Association, pp. 79–99. Clandinin, D. J. and F. M. Connelly. 1998. ‘Personal experience methods’ in N. K. Denzin and Y. S. Lincoln (eds): Collecting and Interpreting Qualitative Materials. Sage, pp. 150–78. Curtis, A. and M. Romney. (eds). 2006. Color, Race, and English Language Teaching: Shades of Meaning. Lawrence Erlbaum. Davies, B. and R. Harre´. 1990. ‘Positioning: The discursive production of selves,’ Journal for the Theory of Social Behaviour 20: 43–63. de Fina, A. 2003. Identity in Narrative: A Study of Immigrant Discourse. John Benjamins. Freeman, M. 2006. ‘Life ‘‘on holiday’’?’ Narrative Inquiry 16/1: 131–8. Ibrahim, A. 1999. ‘Becoming black: Rap and hip-hop, race, gender, identity, and the politics of ESL learning,’ TESOL Quarterly 33: 349–69. Johnson, K. E. and P. R. Golombek. (eds). 2002. Teachers’ Narrative Inquiry as Professional Development. Cambridge University Press. Korobov, N. 2001. ‘Reconciling theory with method: From conversation analysis and critical discourse analysis to positioning analysis,’

Applied Linguistics: 31/2: 301–306 ß Oxford University Press 2010 doi:10.1093/applin/amq004 Advance Access published on 18 February 2010

FORUM

Input Quality Matters: Some Comments on Input Type and Age-Effects in Adult SLA 1

JASON ROTHMAN and 2PEDRO GUIJARRO-FUENTES

1 University of Iowa and 2University of Plymouth E-mails: [email protected]; [email protected]

INTRODUCTION Inspired by Lenneberg’s (1967) and others’ suggestion of a critical period in the domain of primary language acquisition, much research over the past few decades in adult second language acquisition (SLA) has focused on questions pertaining to the extent to which adults retain the cognitive–linguistic capacity to acquire language in the same way as children (see e.g. Birdsong 1999; Long 2005). To date, relatively little research in this general program has acknowledged the significant disparity in several crucial variables that separate naturalistic learners from classroom learners and the literature has much less systematically considered the deterministic role that differences in input type (crucially quality) play (but see e.g. Mun˜oz 2008 and works cited within). In this commentary, we seek to highlight the importance of these dissimilarities, moving beyond obvious disparities in amount of L2 input exposure and different age factors (e.g. chronological age, age of acquisition, age of arrival), focusing instead on possible differences in the grammatical quality of the input classroom learners receive relative to naturalistic learners (L1 and L2 alike). We start by highlighting the impact this scenario can reasonably have on the process of acquisition and on the ultimate attainment of competence outcomes for classroom learners specifically. In doing so, we frame what we consider a more formally complex learning task for classroom L2 learners as compared

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

In accord with the general program of researching factors relating to ultimate attainment and maturational constraints in adult language acquisition, this commentary highlights the importance of input differences in amount, type, and setting between naturalistic and classroom learners of an L2. It is suggested that these variables are often confounded with age factors. Herein, we wish to call attention to the possible deterministic role that the differences in the grammatical quality of classroom input have on development and on competence outcomes. Framing what we see as greater formal complexity of the learning task for classroom learners, we suggest that one might benefit from focusing less on difference and more on how classroom L2 learners, at least some of them, come to acquire all that they do despite crucial qualitative differences in their input.

302 FORUM

A-O-A AND MATURATION For some, the observation that ultimate attainment for adults is quantifiably different, to a greater or lesser extent, in virtually all linguistic domains as compared with child L1 suggests that the underlying mechanisms involved in the two cases of language acquisition are fundamentally different (see Bley-Vroman 2009 and works cited within). The fact that success levels are variable in adult acquisition as opposed to the relative homogeny in L1 acquisition across learners with comparable input exposure has been used to lend support to this position. As Long (2005) and Rothman (2008) rightly point out, the critical period hypothesis (CPH) debate for SLA is not one of do versus do not, it is one of can versus cannot. Observation alone tells us that most adults do not acquire truly native L2 systems, but is it actually the case that they cannot? The latter position is what is claimed by appealing to the maturational effects of a critical (sensitive) period which applies to the normal case of adult SLA. Fortunately, the can versus cannot dichotomy makes strong predictions beyond what is observable from surface production. Among other things, the cannot position predicts that L2 systems will never project beyond what is transferrable from the L1 and what can reasonably be inducted from available input; that is, no incidental acquisition should obtain (absolutely no poverty of the stimulus). To date, there is an ample body of literature that has shown adult L2 learners acquire very subtle properties of the L2 (especially

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

with their naturalistic counterparts and L1 learners of the same target language. Two basic questions dominate current research: (i) who is more likely to become proficient in an L2, an adult or a child? and (ii) how successfully do adults learn an L2? Perfect language mastery is rarely the result for adults and their depressed L2 competence is attributed to some maturational constraints. However, recent research in child L2 acquisition has shown that child L2 is very similar to adult L2 acquisition in developmental sequence (suggesting the important role of L1 transfer) and while L2 acquisition in childhood typically results in better competence outcomes, it is not the case that child L2 inevitably equals L1 acquisition competence either (e.g. see Schwartz 2003; Haznedar and Gavruseva 2008). We take this to suggest that age-of-acquisition (A-o-A), while a seemingly important factor at many levels, might be confounded with other crucial variables that are in fact more deterministic for adult L2. If we are on the right track, and qualitative input difference in L2 acquisition is an explanatory variable for L1/L2 disparity, then it makes sense that child L2 learners would generally be more successful than adults, since most child L2 acquisition, at least the cases studied in recent literature, involves naturalistic L2. That their developmental sequence and, to a lesser extent, their competence is similar to adult L2 (especially naturalistic L2) and different from L1 already tells us that age alone cannot be the (only) variable to explain disparities in route and ultimate attainment.

FORUM

303

WHAT ABOUT THE INPUT? It is well known that L2 learners in a classroom setting, even when the target language is supplemented with some access to natives outside the classroom and through authentic native materials that provide input, necessarily receive considerably less input than those who are acquiring an L2 in a naturalistic setting (and of course native speakers). Since there is nothing that can be done to remedy this imbalance, it is accepted as one of the variables of inherent difference between native and (most) non-native acquisition. However, from non-functionalist, non-frequency points of view on language acquisition, it is

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

complex semantic entailments). Such evidence falsifies the strong L2 critical period position (at least for some domains like syntax and semantics). Since these examined L2 properties are not available from L1 transfer, are not explicitly taught, and are directly absent from available L2 input, it is reasonable to argue that such knowledge obtains via access to internal, in-born linguistic properties (see Slabakova 2008 for review). This has led many researchers to the conclusion that L1/L2 differences cannot be explained under a position that maintains accessibility to the innate language faculty (in part or in whole) as the primary reason for disparity. The A-o-A debate is different from the CPH in SLA on several grounds, although they share many commonalities. The A-o-A position focuses more directly on input. It points out correctly that variables related to age of exposure (including how this is measured) involve differences in the input which is available to the learner, and it does so crucially at periods of time that, relating to the CPH in SLA, are vital for language learning. The intuitive idea that the earlier the exposure takes place, generally, the better the results in ultimate attainment seems to be is supported by the literature, although not as neatly as one might predict (see Long 2005; Rothman 2008). Since all theories of acquisition, despite many paradigmatic differences, agree that input drives acquisition, measuring the timing of significant input exposure (when it is first obtained and occurs in a sustained manner) is crucial and a necessary part of any theory that takes age to be a deterministic factor in a can versus cannot dichotomy. There seems to be a tacit assumption in research methodology that the age-of-first-significant-exposure correlates with the age-of-exposure to native or native-like quality input, irrespective of whether such input is provided in a naturalistic or classroom setting. Since the majority of adult L2 acquisition happens (at least in part) in the classroom and conclusions about fundamental differences are drawn based on the acquisition outcomes of learners whose input was either exclusively or in large part limited to the classroom, this seems like an unfair assumption a priori. In the remainder of this commentary, we will frame why we believe this to be an inaccurate assumption as well as how and why this undermines the conclusions of support for the CPH which have been obtained via studies that employ classroom learners only.

304 FORUM

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

not assumed that less input overall should (must) result in competence outcome differences. The generative paradigm, for example, works under the hypothesis that many linguistic properties are not acquired on the basis of direct evidence from the input, so-called poverty-of-the-stimulus properties, but obtain in grammars (L1 and L2 alike) as a result of linguistic universals that are triggered via the acquisition of properties encoded in the language-specific lexicon of the environment. Under such a scenario, equality in the amount of input would theoretically not be needed for the acquisition of all linguistic properties since what is assumed to be needed is minimally sufficient input that provides the triggers for the system to be configured. Although what is minimally sufficient is itself an ill-defined concept, it seems very plausible that native speakers receive much more input than what would be minimally needed for them to arrive at the grammars they do. So, if L2 learners receive even a fraction of the overall input natives do, and if classroom learners receive much less input than naturalistic L2 learners do, then it is not clear why such unequal input would not provide minimally sufficient amounts for any given property. Of much more concern for the present commentary, and precisely because in our minds it is an important issue which has not received its due attention in SLA literatures, is the question of quality of the input, with respect to differences between naturalistic learners and instructed L2 learners, and the consequences thereof. By quality, we do not intend to infer any type of prescriptive judgements of what can and cannot be said since, strictly speaking, under this interpretation a fair argument could be made that instructed L2 learners receive better quality input. Unlike that of naturalistic learners, it is almost entirely devoid of non-standard input. Alternatively, by quality, we refer to the grammatical accuracy of the input provided (e.g. the syntactic, semantic, and morphological forms). Much research has shown that L2 learners, even those at very high levels of L2 proficiency, continue to make errors in their L2 speech, especially at the level of functional morphology (see Slabakova 2008; Lardiere 2009 for a review). Of course, errors at all levels are much more abundant at low levels of proficiency. With this in mind, we must ask ourselves: what are the primary sources of aural input available to instructed L2 learners, and is this the same for naturalistic L2 learners? Instructed L2 learners are often, although not exclusively so, in learning environments where the teachers are L2 learners themselves. Outside of the Western world and even within, depending on the language being taught as an L2, this scenario is the norm. Even if some of the instructors they have had over the course of the many years of L2 learning are natives, it is likely that some are not. Since even highly proficient (highly advanced) L2 learners show variability and optionality in use (regardless of whether or not psycholinguistic studies convincingly demonstrate that in such populations the errors are more performance issues than reflective of competence), one can safely assume that the input they receive is qualitatively different (at least on the surface) from native children and naturalistic adults who find themselves

FORUM

305

CONCLUSION We wish to make it clear that we in no way intend to suggest that quantitative or qualitative differences in input alone suffice to explain all L1/L2 disparities. Alternatively, this commentary’s goal was to bring to the surface two interrelated points: (i) the expectation that L2 competence based on input which comes primarily from instructional settings would be (or could be) comparable to native controls or even naturalistic non-natives constitutes a comparative fallacy which is based, among many other variables, on qualitative input differences and (ii) that the A-o-A debate does not take (i) into consideration to the extent that it should. With that said, we do not intend to deny that age confers differences and that, at the onset of acquisition, it seems to result in less accurate competence and performance. Future research that acknowledges the extra-burdened learning task of classroom learners, who must sift through input that itself provides non-target forms and target forms, and/or seeks to quantify the extent that this actually matters, will provide us with much needed information. More studies which focus both on developmental sequence and ultimate attainment outcomes and compare naturalistic versus non-naturalistic L2 learners are needed, especially when age can truly isolate as an independent variable from qualitative differences in input by adding child L2 in naturalistic and non-naturalistic settings. Such studies might show that the amount of input and especially its quality are more deterministic factors than mere age in adult language acquisition. In light of input quality

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

in a vastly different linguistic environment. Moreover, L2 learners receive constant feedback from other L2 learners in the classroom setting and thus are provided with L2 input that is sure to be inconsistent with morphological and syntactic structures; this makes the task of determining what the native grammatical properties are more complex. In this sense, there is a competition within the input provided that natives and naturalistic learners do not need to sift through. In light of the input that the L2 classroom learners receive, it is perhaps not surprising that developmental sequence as well as ultimate attainment is different and that the quality of the input issue is one of the many variables that will ultimately help to explain (some) L1/L2 differences. The fact that learners who continue to demonstrate variability and optionality in their L2 performance even at high levels of proficiency have been shown to possess sophisticated knowledge of very subtle L2 properties (especially semantic entailments of new L2 properties, see Slabakova 2008) is thus a testament to the fact that L2 acquisition is not (as) fundamentally different in potential and process as some theories of L2 acquisition propose. Qualitative differences in input might serve to explain some of the asymmetries in the outcomes of classroom versus naturalistic L2 learning and the fact that there are similarities as well can be used as evidence which points in the direction that the processes themselves are not fated to be different, at least not on the so-called linguistic maturational grounds.

306 FORUM

differences and the noise such creates for classroom learners, the success that some of these learners show in ultimate attainment becomes even more remarkable. A focus on these successes despite a task that is seemingly stacked against them in crucial ways, we believe, not only provides motivation for shifts in the A-o-A debate, but also highlights the latent potential that adults have for language learning in general despite anecdotal evidence to the contrary.

REFERENCES hypothesis.’ International Review of Applied Linguistics 43/4: 287–317. Mun˜oz, C. 2008. ‘Symmetries and asymmetries of age effects in naturalistic and instructed L2 learning.’ Applied Linguistics 29/4: 578–96. Rothman, J. 2008. ‘Why not all counter-evidence to the critical period hypothesis is equal or problematic: Implications for SLA.’ Language and Linguistics Compass 2/6: 1063–88. Schwartz, B. 2003. ‘Child L2 acquisition: Paving the way’ in Beachly B., A. Brown, and F. Conlin (eds): BUCLD 27: Proceedings of the 27th Annual Boston University Conference on Language Development. Cascadilla Press, pp. 26–50. Slabakova, R. 2008. Meaning in the Second Language. DeGruyter.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Bley-Vroman, R. 2009. ‘The evolving context of the fundamental difference Hypothesis.’ Studies in Second Language Acquisition 31/2: 175–98. Birdsong, D. (ed.) 1999. Second Language Acquisition and the Critical Period Hypothesis. Lawrence Earlbaum Associates. Haznedar, B. and E. Gavruseva. (eds) 2008. Current Trends in Child Second Language Acquisition: A Generative Perspective. Benjamins. Lardiere, D. 2009. ‘Some thoughts on the constructive analysis of features in second language acquisition.’ Second Language Research 25/2: 173–227. Lenneberg, E. 1967. Biological Foundations of Language. Wiley. Long, M. 2005. ‘Problems with supposed counter-evidence to the critical period

Applied Linguistics 31/2: 307–319

ß Oxford University Press 2010

REVIEWS Hannes Kniffka: WORKING IN LANGUAGE AND LAW: A GERMAN PERSPECTIVE. Palgrave Macmillan, 2007.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

This volume brings together the life’s work of Hannes Kniffka with two principal objectives: first, to provide a coherent collection of this important body of work; and second, to provide this collection in English. Kniffka’s work to date has mainly appeared in German. One implication, although not explicitly stated in the book, is that the English-speaking work on forensic linguistics has not always been cognisant of work published in German, and in fact the author claims that the work in Germany was already well advanced when it was only beginning in English-speaking contexts. The book reminds us yet again of the importance of those who work in linguistics to also practice linguistics in the sense of knowing and using different languages and being able to access sources in a variety of languages. As Roger Shuy states in the preface, the book ‘opens an important door to those unfamiliar with the German language and context, and it helps them compare and learn from the theory and practice of the earlier and more recent work by Kniffka, who founded and has practiced forensic linguistics in Germany since the 1970s’ (p. xii). The book is divided into three parts. Part 1, titled ‘The Interdisciplinary Status of Forensic Linguistics’ is restricted to an overview chapter, which gives a very comprehensive introduction to the area of forensic linguistics, differentiating it from forensic phonetics, forensic handwriting analysis, and document analysis, and situating forensic linguistics ‘as a branch of applied linguistics and as one of the auxiliary forensic sciences’ (p. 16). In Part 2, the author moves on to look more specifically at his own body of work carried out within the particular context of the German legal system. The first chapter is a translation of an excerpt from a longer paper written in 1981. The main focus of linguists’ work was in attempting to attribute authorship to anonymously written documents, which are presented as evidence, e.g. ransom notes, libellous statements, etc. One key point, which emerges, is that there is no magic formula to find ‘linguistic fingerprints’ in a document, because, as Kniffka establishes fairly exhaustively, there is really no such thing. Forensic linguistics can never have the same scientific certainty as e.g. DNA analysis, because we are dealing with humans using language in various contexts and with all sorts of unpredictable variables at play. The second chapter in Part 2 ‘Status and Tasks of Forensic Linguistic Authorship Analysis’ contains a translation of a paper, which was primarily written for the police—in an attempt to explain the work of forensic linguists. This particular chapter provides evidence of Kniffka’s own firm belief, which he restates many times in the book, that linguists need to be able to explain what they do to nonlinguists; that they need to analyse their own language, not just other people’s

308 REVIEWS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

language. The final chapter in this part does not have an explicitly forensic linguistics focus, but instead looks at shibboleths as ‘data of linguistic behaviour’, an issue that, Kniffka argues convincingly, is helpful in attributing authorship. Part 3 of the book attempts an overview of forensic linguistics research in Germany, particularly focusing on the author’s work, and taking the particularities of the German legal system into account. The first chapter, thus, explains the features of German defamation laws and the work of the linguist in interpreting them through the forensic analysis of documents. Again, in this chapter, Kniffka stresses the importance of the interaction between linguist and lawyer, and the need for both to understand each other and each other’s roles. The second chapter, ‘A Heuristic Author and Writer/Typer Taxonomy’, deals with the issue of establishing ‘authorship’ of a text versus ‘writership’ and ‘typership’ of a text, and exemplifies, using German cases, a framework for explaining the many possible variations between the two and the most frequently found combinations. The last two chapters in Part 3 focus on the role of orthographic data in forensic linguistics, particularly in the area of author attribution in Germany. Orthographic data, Kniffka tells us, ‘can never function as sufficient evidence alone, but frequently function as evidence of data of written texts indirectly relevant for authorship attribution’ (p. 22). Part 4, the final part of the book, is devoted to ‘Outlook and Future Tasks’. The first of the two concluding chapters is ‘in short, a summary of the real-life experience the author has gained in some 35 years of giving expert testimony in and for German courts’ (p. 23), and the objective of the chapter is to show the misunderstandings between the judiciary and linguists, in the hope of highlighting research that needs to be carried out to bridge the gap between these disciplines. In the final chapter, the author gives his outlook for forensic linguistics research, with a particular call for interdisciplinary and international cooperation. The future seems to lie with corpus-driven approaches and computer-aided quantitative analysis, since, in the author’s own words ‘everything that can be measured should be measured’ (p. 267). As someone quite new to the area of forensic linguistics and the whole area of how linguistics interacts with and is applied in the law, I found the book genuinely fascinating and insightful. A good many issues of concern to all linguists are explored in the volume, e.g. the issue of what exactly a text is and what it is not. In Kniffka’s words, ‘[. . .] texts are frozen products of linguistic behaviour and have to be described as such. They do not show a total and in-depth representation of the linguistic behaviour involved’ (p. 68). The book is written in a no-nonsense sort of style, which is refreshing. I also enjoyed the use of German and translations in the book, and the author’s clear concern with issues of translation. The account is in some ways quite personal, Kniffka giving a ‘biography’ of each of the texts appearing in the collection, but in other ways this reflects a concern with detail, objectivity and

REVIEWS 309

Reviewed by Helen Kelly-Holmes University of Limerick, Ireland E-mail: [email protected] doi:10.1093/applin/amq003 Advance Access published on 22 February 2010

Nancy H. Hornberger (ed.): CAN SCHOOLS SAVE INDIGENOUS LANGUAGES? POLICY AND PRACTICE ON FOUR CONTINENTS. Palgrave Macmillan, 2008. This book deals with pedagogical and political issues in implementing Indigenous1-language education. The four main chapters originated at a colloquium at the AILA/AAAL conference in 2005, and are supplemented by commentaries from wider contexts. Vuokko Hirvonen’s chapter on Sa´mi education in Norway introduces several themes which run through the book, especially the advantages of mother-tongue education, and the need for Indigenous control of the curriculum. Despite considerable progress, Sa´mi schools do not yet attain the goal of full bilingualism due to insufficient quality and quantity of input, lack of teachers able to teach both language and content, inadequate materials, resources, and assessment. These lead one to question the commitment of

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

a scientific approach, which I think must be an influence from the author’s practice of forensic linguistics. Nonetheless, I would have liked to know why or how Kniffka became interested in and began to develop this area in a pioneering way, and why it seems to have developed in Germany in this particular way. Is it something to do with the particular linguistic and/or legal culture? The actual cases and data that have been gathered and analysed over 35 years are fascinating: there is the case of a fake edition of a company newsletter put together by a disgruntled former editor; and that of Ms Y, who filed an action against Mr A for his use of the term ‘Konkubine’ towards her; there is the fact that extortion letters tend to be less than 200 words in length with no marked or excessive use of punctuation, while defamation letters tend to be between 200 and 400 words with highly marked use of punctuation. The structure of the book is, however, not always logical, and, ultimately, despite a very good linking chapter at the beginning, my impression is that of a very good anthology of collected papers rather than a coherent book from start to finish. Each of the chapters is self-contained and can be read on its own. Read in this way, the volume has a great deal to teach us about working with language and law, and beyond this, about working with language in any applied field.

REVIEWS 309

Reviewed by Helen Kelly-Holmes University of Limerick, Ireland E-mail: [email protected] doi:10.1093/applin/amq003 Advance Access published on 22 February 2010

Nancy H. Hornberger (ed.): CAN SCHOOLS SAVE INDIGENOUS LANGUAGES? POLICY AND PRACTICE ON FOUR CONTINENTS. Palgrave Macmillan, 2008. This book deals with pedagogical and political issues in implementing Indigenous1-language education. The four main chapters originated at a colloquium at the AILA/AAAL conference in 2005, and are supplemented by commentaries from wider contexts. Vuokko Hirvonen’s chapter on Sa´mi education in Norway introduces several themes which run through the book, especially the advantages of mother-tongue education, and the need for Indigenous control of the curriculum. Despite considerable progress, Sa´mi schools do not yet attain the goal of full bilingualism due to insufficient quality and quantity of input, lack of teachers able to teach both language and content, inadequate materials, resources, and assessment. These lead one to question the commitment of

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

a scientific approach, which I think must be an influence from the author’s practice of forensic linguistics. Nonetheless, I would have liked to know why or how Kniffka became interested in and began to develop this area in a pioneering way, and why it seems to have developed in Germany in this particular way. Is it something to do with the particular linguistic and/or legal culture? The actual cases and data that have been gathered and analysed over 35 years are fascinating: there is the case of a fake edition of a company newsletter put together by a disgruntled former editor; and that of Ms Y, who filed an action against Mr A for his use of the term ‘Konkubine’ towards her; there is the fact that extortion letters tend to be less than 200 words in length with no marked or excessive use of punctuation, while defamation letters tend to be between 200 and 400 words with highly marked use of punctuation. The structure of the book is, however, not always logical, and, ultimately, despite a very good linking chapter at the beginning, my impression is that of a very good anthology of collected papers rather than a coherent book from start to finish. Each of the chapters is self-contained and can be read on its own. Read in this way, the volume has a great deal to teach us about working with language and law, and beyond this, about working with language in any applied field.

310 REVIEWS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

education authorities, so it is sobering to remember that Norway is now the most progressive state with regard to the recognition of Sa´mi languages, culture, and political rights. For this chapter, it would have been helpful to include more illustrations of classroom practices and teachers’, students’, and parents’ views. The relationship between school and language revitalization is mentioned only tantalisingly at the very end. In his chapter on Latin America, Luis Enrique Lo´pez forcefully makes the point that Indigenous schools are not only about language, but also the revalorization of Indigenous culture and the empowerment of communities. If curricula are controlled by the government, unsurprisingly schools may not fulfil these goals. Lo´pez refers to debates which counterpoise minority rights and national unity: ‘intercultural bilingual education is criticised for its high costs, its insufficient gains in terms of school achievement . . . or even for intensifying racism and promoting opposition and division within the nation’ (p. 45). Although it can be argued that social cohesion cannot be achieved while a large sector of the population is disadvantaged, such arguments have wide currency2 and need to be countered before the benefits of bilingual education and minority empowerment are accepted. In their chapter on Ma¯ori-medium education in Aotearoa/New Zealand, Stephen May and Richard Hill point out that there is little information on the educational effectiveness of the programmes (p. 68), although they are held up as a model of language revitalization. May and Hill note that ‘even in Indigenous language programs that have traditionally drawn primarily from L1 speakers, more L2 speakers are increasingly present’ (p. 70). They cite the case of Navajo education in Arizona, ‘one of the strongest and longest established in the USA—only 50 per cent of Navajo now speak their own language and their numbers are declining each year’. Although May and Hill do not say so, this must surely call into question the effectiveness of such programmes for language maintenance. The authors propose that further development should focus on consolidating programmes which provide the highest proportion of exposure to Ma¯ori, which are found to be the most effective; but as elsewhere, the unavailability of fluent and well-trained teachers causes problems. It is clear that there is fundamental misunderstanding of the processes and benefits of bilingual education among parents. Even supportive parents are concerned that children will not gain sufficient proficiency in English, and withdraw them after a few years. Students gain conversational competence in Ma¯ori relatively quickly, but academic language proficiency takes longer; they therefore do not reach a ‘sufficient bilingual threshold to be able to transfer literacy skills effectively from one language to the other’ (p. 90). It is to be hoped that this book will help counter such misunderstandings, although its audience is more likely to be academic than popular. Nicanor Rebolledo Recendiz describes attempts to halt language shift when Indigenous people migrate to urban areas. The chapter describes the Hn˜a¨hn˜o¨

REVIEWS 311

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

speakers’ struggle to maintain their way of life in Mexico City. Discrimination and distrust lead to low participation and achievement in the public school system, which aims for cultural uniformity and has no provision for bilingual or multicultural education. The chapter describes a project which promotes ‘the recognition of the reality of Indians within the public schools . . . in an approach which tries to recover Indian cultures and languages, [and] recognizes the value of diversity’ (p. 106). The chapter was written shortly after the start of the project in 2004; it is a pity that developments since are not evaluated. The chapter describes problems which are common to many contexts: (i) a complete lack of Hn˜a¨hn˜o¨ speakers in the teaching profession, (ii) even speakers do not value Hn˜a¨hn˜o¨, (iii) there is no standard written form, and (iv) there are no materials. Part II of the book consists of ‘International perspectives on the case studies’. Leena Huss draws parallels between the Sa´mi and Ma¯ori situations, e.g. withdrawal of children and the lack of programmes targeting students with ‘weak or no competence’ (p. 127). As in Mexico, Indigenous people are increasingly leaving traditional lands to seek work in cities, while Sa´mi education continues to focus on ‘authenticity’. Huss concludes on a positive note, however: ‘the struggle for revitalisation . . . is still for many individuals and groups a rehabilitating, healing and empowering experience’ (p. 133). Nkonko M. Kamwangamalu summarizes debates concerning mother-tongue education in Africa, pointing out that in Western (and some Asian) countries, mother-tongue education is seen as enabling and empowering; but in Africa it is frequently perceived (and experienced) as the opposite (p. 141). The debate in Africa has focused on educational achievement, literacy, and political empowerment rather than language revitalization; as Kamwangamalu notes, endangered language revitalization is relatively rare in Africa, while linguistic rights are seen as ‘a matter of small concern . . . when a person is fully pre-occupied with making a living and doing what it takes to survive’. Where ‘the instrumental value of a language weighs far more than its cultural value’ (p. 149), language shift and loss will continue. Bernard Spolsky is the only contributor who addresses an issue that he calls ‘riding the tiger’: the risk of relying on institutions which formerly (and so effectively) suppressed Indigenous languages and cultures. He also problematizes the term ‘Indigenous’: ‘Who gets to use the label? After all, the Spanish arrived in the south-western USA before the Navajo, and African languages seem to have moved around a great deal, and their present distribution was not set at the time of colonization’ (p. 153). Stigmatization and language shift are common to both ‘Indigenous’ and ‘immigrant’ languages (Edwards 2000; Extra and Gorter 2001). I would have hoped for more discussion of such issues. Teresa McCarty cites ‘strong’ forms of bilingual education (over 50% Indigenous-language medium) and Hawaiian-medium pre-schools and elementary schools as success stories. Other beneficial effects cited include: elevating the status of an Indigenous language, the promotion of Indigenous

312 REVIEWS

Reviewed by Julia Sallabank School of Oriental and African Studies, London, UK E-mail: [email protected] doi:10.1093/applin/amq005 Advance Access published on 21 February 2010

NOTES 1 I will follow the book’s practice of writing ‘Indigenous’ with a capital ‘I’ ‘in keeping with the practice of Indigenous organizations’. 2 At a colloquium on multilingualism at

the British Museum in December 2008, the Chair of the UK Equality and Human Rights Commission, Trevor Phillips, expressed a not dissimilar viewpoint.

REFERENCES Edwards, V. 2000. ‘‘Community Languages’’ in Price G. (ed.): Languages in Britain and Ireland. Blackwell, pp. 213–28.

Extra, G. and D. Gorter. 2001. The Other Languages of Europe: Demographic, Sociolinguistic and Educational Perspectives. Multilingual Matters.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

rights, ethnolinguistic pride, strengthened intergenerational ties, orthographic standardization, the training of Indigenous teachers, and the involvement of parents, ‘many of whom become speakers themselves’. However, these programmes rely on government funding; if it were to be withdrawn, would the languages decline again, given that they have not re-established themselves in the family? As noted earlier, not all the benefits have yet been fully reaped; and ‘not all native Americans believe that it is appropriate or desirable to teach Indigenous languages in the alien environment of the school, or to instruct non-tribal members in tribal languages’ (pp. 172–3). McCarty’s chapter, which acts as a conclusion, is the only one to respond directly to the question in the book’s title. She sums up that ‘schools alone cannot do the job, but in tandem with other institutions, they can be (and have been) a strategic resource for exerting Indigenous language and education rights’ (p. 175). This book makes a considerable contribution to knowledge about the processes and outcomes of Indigenous-language education. However, there are several controversial issues which are barely addressed: standardization and the ‘updating’ of languages, the definition of (and boundaries between) languages and groups in increasingly pluralistic societies, and a comparison with other types of language revitalization measure. I was disappointed not to see discussion of these, while other matters such as community control and degrees of immersion were discussed in several chapters. I would also have liked to see more synthesis of the findings, especially lessons that might be learned by other communities. Despite these shortcomings, this book contains much that is useful. I hope that it will inspire more evaluation of the role of schools in Indigenous and endangered language revitalization.

REVIEWS 313

E. Alco´n Soler and Alicia Martı´nez Flor (Eds): INVESTIGATING PRAGMATICS IN FOREIGN LANGUAGE LEARNING, TEACHING AND TESTING. Multilingual Matters, 2008.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

In the last 10 years, the field of interlanguage pragmatics has seen a marked increase in the number of edited volumes (e.g. Rose and Kasper 2001; Bardovi-Harlig et al. 2006; Puetz and Neff-van Aertselaer 2008) and special issues (e.g. Alco´n Soler and Martı´nez Flor 2005; Barron and Warga 2007) being published that provide insights into how pragmatics may be learned and taught. The volume under review with its focus on pragmatics in the foreign language context is a welcome addition to the literature available. Following a preface by Amy Snyder Ohta, and an introductory chapter outlining the rationale for the volume by the two editors, the book is divided into three parts, with foci on learning (Part 1), teaching (Part 2), and testing (Part 3). In the second chapter entitled ‘Language acquisition theory and the acquisition of pragmatics in the foreign language classroom’, Margaret A. DuFon calls for a theoretical approach in second language acquisition that is of an integrative nature and regards social and mental aspects as equally important. She argues that the interactionist and holistic nature of language socialization theory offer researchers the opportunity to explore pragmatic acquisition in a more integrated paradigm. Also advocating a social-interactional participation framework, the authors of the following chapter, Yumiko Tateyama and Gabriele Kasper, investigate foreign language learners’ exposure to and participation in request utterances in their study entitled ‘Talking with a classroom guest: opportunities for learning Japanese pragmatics’. The results of their investigation show that the presence of the classroom guest helped expose foreign language learners to a greater variety of request strategies than they would normally encounter in a typical foreign language classroom context. In contrast to the two preceding chapters, Tim Hassall’s study entitled ‘Pragmatic performance: what are learners thinking’ uses a cognitive framework to investigate the use of address and self-reference terms by Australian learners of Indonesian. Data of elicited requests and complaints were collected with role plays and retrospective verbal protocols. The results of the study reveal that learners with lower proficiency concentrated more on linguistic planning (e.g. searching for lexical items) than on pragmatic issues, whereas the learners with higher proficiency reflected more on the latter. Regarding learners’ production of requests and complaints, the investigation suggests that learners’ use of a particular form does not automatically indicate firm control over it. In Chapter 5, Tarja Nikula examines Finnish secondary school students’ use of English in biology and physics lessons in her qualitative study entitled ‘Learning pragmatics in content-based classrooms’. The teachers in

314 REVIEWS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

this investigation were Finnish native speakers who did not have formal qualifications in English. The framework is broadly sociocultural and the focus is on the students’ ability to negotiate potentially face-threatening situations such as misunderstandings and disagreements. The results of the study indicate that the students did not employ a wide range of discourse markers and pragmatic particles. While they generally seemed to be able to negotiate face-threatening situations in a manner that did not cause their interlocutors offense, it is not clear whether this is because all of the interlocutors shared the pragmatic norms of their first language and therefore might not have taken offense where native speakers of other languages might have done so. Also, as the author focuses on the students’ interactions, the extent to which subject teachers with no English qualifications can help learners improve their pragmatic competence is unfortunately not addressed in great detail. In a case study presented in Chapter 6 entitled ‘Computer-mediated learning of L2 pragmatics’, Marta Gonza´lez-Lloret investigates over a 10-week period how collaborative electronic exchange affects foreign language learners’ ability to attend to addressivity in Spanish. Using a conversation analytical approach, the author shows that her learners’ use of address terms and pronouns changes from ‘chaotic variation to an exclusive use of the informal, the expert norm established by the Spanish speaker’ (p. 126) which seemed to be the result of the explicit and repeated corrections provided by the native speaker interlocutor in the computer-mediated context. Chapter 7 entitled ‘Using translation to improve pragmatic competence’ by Juliane House is the first of three chapters that focus on pragmatics and language teaching. In her article, House argues for a reassessment of the role of translations in language teaching. She posits that translations can usefully be employed to raise advanced language learners’ awareness of pragmatic and cultural differences. In Chapter 8 entitled ‘Effects on pragmatic development through awareness raising instruction: Refusals by Japanese EFL learners’, Sachiko Kondo shows how after 12 weeks of pragmatics instruction L2 learners’ performance of refusal strategies became more native-like. Exercises in which learners compared typical refusal strategies in their native and target language resulted in learners employing more positive strategies, such as expressing gratitude and future acceptance, whereas learners had mainly relied on showing regret prior to instruction. The study reported on in the following chapter by Zohreh R. Eslami and Abbass Eslami-Rasekh entitled ‘Enhancing the pragmatic competence of non-native English-speaking teacher candidates (NNESTCs) in an EFL context’ shares some similarities with Kondo’s in that it was conducted over a similar length of time, 14 weeks, and focuses on speech acts. The results of their investigation are equally encouraging as Kondo’s. Pre- and post-instruction data show that using a variety of teaching approaches, such as exposing students to research findings and giving them opportunities to conduct their own pragmatic research, leads to a significant increase in their pragmatic competence.

REVIEWS 315

Reviewed by Gila A. Schauer Lancaster University, UK E-mail: [email protected] doi:10.1093/applin/amq006 Advance Access published on 21 February 2010

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Chapter 10 by Sayoko Yamashita entitled ‘Investigating interlanguage pragmatic ability: what are we testing’ is the first of three chapters examining pragmatics and testing. Yamashita argues that testing in interlanguage pragmatics has primarily focused on exploring learners’ ability to produce and comprehend speech acts. She asserts that ‘testing only a limited area does not represent or describe a whole picture of the learners’ pragmatic ability. (. . .) We should expand our research agenda to understand the construction of pragmatics in more depth’ (p. 219). In Chapter 11 entitled ‘Raters, functions, item types and the dependability of L2 pragmatics tests’, James Dean Brown examines the reliability of commonly used data elicitation techniques: written and oral discourse completion tasks (DCTs), multiple choice tasks, role plays, and self-assessments. On the basis of a review of previous studies that had compared the reliability of these techniques, he concludes that the multiple choice test is the only instrument that does not achieve a satisfactory reliability score in the studies surveyed. All other techniques tended to generate moderate to high reliability scores. In the final chapter of the volume entitled ‘Rater, item and candidate effects in discourse completion tests: A FACETS approach’, Carsten Roever uses Many Facet Rasch Measurement to reanalyze data from his 2005 study, which was based on DCT results and focused on learners’ pragmalinguistic competence. The findings of his investigation show that the test was very reliable overall, which Rover concludes is a very positive result for studies using DCTs as a data collection instrument. One of the strengths of this edited volume is that it brings together a wide and diverse range of papers. The chapters in the first part of the book provide a rich resource for anyone wishing to obtain insights into theoretical approaches in the area of pragmatics and language learning. Although all articles in this part make a valuable contribution, the impact of this section would have been even greater if a theoretical article on cognitive theories had been added. House’s thought-provoking paper on translation and pragmatics could initiate discussions about a re-evaluation of the role of translation in language teaching, while the other two articles in this part provide encouraging results for proponents of pragmatics instruction in the foreign language context. The final three papers on testing raise a variety of issues concerning elicited data that will be of interest to researchers and advanced students who are in the process of developing their own research instruments. Overall, this is a timely and stimulating volume that will be a valuable addition to the bookshelves of researchers, teacher trainers, teachers, and advanced students interested in interlanguage pragmatics.

316 REVIEWS

REFERENCES Alcon Soler, E. and A. Martinez Flor. (eds) 2005. ‘Pragmatics in instructed language learning’ (Special issue) System, 33/3. Bardovi-Harlig, K., C. Felix-Brasdefer, and A. Omar. (eds) 2006. Pragmatics and Language Learning. Second Language Teaching and Curriculum Center University of Hawaii.

Barron, A. and M. Warga. (eds) 2007. ‘Special Issue in Acquisitional Pragmatics’ Intercultural Pragmatics, 4/2. Puetz, M. and J. Neff-van Aertselaer. (eds) 2008. Developing Contrastive Pragmatics: Interlanguage and Cross-Cultural Perspectives. Mouton de Gruyter, pp. 399–426. Rose, K. R. and G. Kasper. (eds) 2001. Pragmatics in Language Teaching. Cambridge University Press.

This work offers an account of the emergence and development of pidgins and creoles. The intended readership consists of specialists in language contact, in general linguistics and in second language acquisition (SLA). Chapter 1 starts with the definitions of several types of contact languages: pidgins, creoles, indigenized varieties, and language shift varieties. Next, the author states the research questions that the book aims to answer. The chapter ends with an outline of the structure of the book. Chapter 2 looks at the morphological simplicity in pidgins. This is illustrated with examples from Pidgin Fijian, and discussed in the light of the various definitions and suggestions for its measurement. The author proposes as an indicator of morphological simplicity a scale from ‘lexicality’ to ‘grammaticality’: content item > grammatical item > clitic > inflectional affix (p. 23). The morphological simplicity of restricted pidgins is shown to be similar to that found in naturalistic adult SLA. The various accounts of simplicity in SLA research (based on universal grammar, constructivist or speech production models) are critically evaluated. The stabilization of pidgins is attributed to mixing and levelling. The retention of morphological simplicity is the outcome of ‘strategic SLA’, that is a process of learning just enough to communicate. Morphological simplicity and expansion in creoles are examined in Chapter 3. On the basis of examples from Bislama, creoles are shown not to be simpler overall than other languages, including their lexifiers. The process of morphological expansion is illustrated with rich comparative data from early and expanded Hawai’i Pidgin English. Chapter 4 analyses the possible sources of morphological expansion: language-internal developments, linguistic universals, other languages (the lexifier, other pidgins, and creoles, the substrate languages). Particular attention is paid to the Language Bioprogram Hypothesis, whose main tenets are disconfirmed by the analysis of the development of Hawai’i Creole. The examination of examples of substrate influence in Melanesian

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Jeff Siegel: THE EMERGENCE OF PIDGIN AND CREOLE LANGUAGES. Oxford University Press, 2008.

316 REVIEWS

REFERENCES Alcon Soler, E. and A. Martinez Flor. (eds) 2005. ‘Pragmatics in instructed language learning’ (Special issue) System, 33/3. Bardovi-Harlig, K., C. Felix-Brasdefer, and A. Omar. (eds) 2006. Pragmatics and Language Learning. Second Language Teaching and Curriculum Center University of Hawaii.

Barron, A. and M. Warga. (eds) 2007. ‘Special Issue in Acquisitional Pragmatics’ Intercultural Pragmatics, 4/2. Puetz, M. and J. Neff-van Aertselaer. (eds) 2008. Developing Contrastive Pragmatics: Interlanguage and Cross-Cultural Perspectives. Mouton de Gruyter, pp. 399–426. Rose, K. R. and G. Kasper. (eds) 2001. Pragmatics in Language Teaching. Cambridge University Press.

This work offers an account of the emergence and development of pidgins and creoles. The intended readership consists of specialists in language contact, in general linguistics and in second language acquisition (SLA). Chapter 1 starts with the definitions of several types of contact languages: pidgins, creoles, indigenized varieties, and language shift varieties. Next, the author states the research questions that the book aims to answer. The chapter ends with an outline of the structure of the book. Chapter 2 looks at the morphological simplicity in pidgins. This is illustrated with examples from Pidgin Fijian, and discussed in the light of the various definitions and suggestions for its measurement. The author proposes as an indicator of morphological simplicity a scale from ‘lexicality’ to ‘grammaticality’: content item > grammatical item > clitic > inflectional affix (p. 23). The morphological simplicity of restricted pidgins is shown to be similar to that found in naturalistic adult SLA. The various accounts of simplicity in SLA research (based on universal grammar, constructivist or speech production models) are critically evaluated. The stabilization of pidgins is attributed to mixing and levelling. The retention of morphological simplicity is the outcome of ‘strategic SLA’, that is a process of learning just enough to communicate. Morphological simplicity and expansion in creoles are examined in Chapter 3. On the basis of examples from Bislama, creoles are shown not to be simpler overall than other languages, including their lexifiers. The process of morphological expansion is illustrated with rich comparative data from early and expanded Hawai’i Pidgin English. Chapter 4 analyses the possible sources of morphological expansion: language-internal developments, linguistic universals, other languages (the lexifier, other pidgins, and creoles, the substrate languages). Particular attention is paid to the Language Bioprogram Hypothesis, whose main tenets are disconfirmed by the analysis of the development of Hawai’i Creole. The examination of examples of substrate influence in Melanesian

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Jeff Siegel: THE EMERGENCE OF PIDGIN AND CREOLE LANGUAGES. Oxford University Press, 2008.

REVIEWS 317

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Pidgin and Hawai’i Creole leads the author to the conclusion that the use of lexical items from the lexifier, with the grammatical function of morphemes from the substrate languages, is the most important factor in morphological expansion. Chapter 5 focuses on word order transfer and functional transfer; the latter is defined as the use of L2 lexical items with L1 grammatical functions (p. 108). After examining transfer in SLA, bilingual foreign language acquisition, and second language use, the author concludes that functional transfer typically occurs in the latter context. On this view, called ‘Compensatory Transfer’ (p. 133), lexical items from the lexifier are used in expanded pidgins and creoles with grammatical functions of morphemes from the substrate languages, to compensate for the inadequate resources of the restricted pidgin predecessor. The Compensatory Transfer account is convincingly shown to fare better than alternative accounts such as the Language Bioprogram Hypothesis, relexification, contact-induced grammaticalization, functional interference, or the Two Targets model. Chapter 6 examines two types of constraints on the influence of substrates: availability constraints, which account for the transfer of features, and reinforcement principles, which influence the retention of only particular features. Both availability and retention are determined by sociolinguistic and linguistic factors. Concerning the sociolinguistic factors, the author proposes a ‘Shifter Principle’. According to this principle, the substrate languages whose speakers first shift to an expanding pidgin are most likely to provide features that are transferred (p. 151). The linguistic factors discussed are markedness, perceptual salience, transparency, simplicity, frequency, and congruence. Siegel demonstrates that congruence and perceptual salience are the most significant factors in the transfer of features, while frequency plays the most important part in their retention. Substrate reinforcement is the topic of Chapter 7. The author analyses five grammatical features which differentiate between the modern varieties of Melanesian Pidgin: demonstratives, the adjectival and numeral marker -pela/-fala, relative clause marking, progressive marking, and modality marking. These differences are convincingly attributed to reinforcement by the different substrate languages of Tok Pisin (Papua New Guinea), Bislama (Vanuatu), and Pijin (Solomon Islands). Chapter 8 demonstrates the predictive power of the availability constraints and of the substrate reinforcement principles by means of two case studies: the tense–modality–aspect system of the French-lexified creole Tayo (New Caledonia) and the verb phrase of the English-lexified creole Roper Kriol (Australia). In Chapter 9, the author looks at decreolization and the creole continuum. First, compelling evidence is adduced against the alleged existence of a post-creole continuum between the varieties of Melanesian Pidgin and English. This is taken to support the view that a creole continuum is not the result of decreolization but of variation present in the creole throughout its history. Next, decreolization in Hawai’i Creole is examined. In addition to conventional decreolization (i.e. the replacement of creole forms with forms from the lexifier), Siegel demonstrates the existence of what he calls ‘covert decreolization’, defined as the change in

318 REVIEWS

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

the function or use of creole forms in the direction of the lexifier (p. 259). Covert decreolization in Hawai’i Creole is illustrated with examples of more English-like use of articles, the past tense marker, the copula, and the completive aspect marker. Chapter 10 summarizes the findings and discusses their implications for several competing theories: the superstratist account, contact-induced grammaticalization, relexification, the Two Targets model, and the Language Bioprogram Hypothesis. I would like to make a few minor comments. The author states that ‘[t]his book deals primarily with contact varieties that have emerged in the Australia-Pacific region’ (p. 1). This is an understatement, since the pidgins and creoles analysed are exclusively Pacific ones, while Atlantic pidgins and creoles are only briefly referred to. Phonology is conspicuously absent, except for some extremely brief discussions (pp. 241, 244–5 and 247), although the emergence and development of the inventory of segments and of syllable types in pidgins and creoles would have been worth examining. Siegel writes that the Language Bioprogram Hypothesis ‘does not account for the sources of expansion in the cases of features not purported to be part of the bioprogram’ (p. 68). This criticism is unwarranted since a theory cannot be blamed for not explaining what it does not aim to explain. The transitive suffix occurring in Melanesian Pidgin is derived from English him or them (p. 87). While him is the generally accepted etymon, no evidence is presented for them as a second possible source. The variants of third person pronouns used in earlier, unstable Melanesian Pidgin listed by the author (p. 176) include ’e and i, which are just different transcriptions of what was presumably [i]. The apostrophe in the English-influenced transcription ’e notes the absence of [h], whereas i reflects the French-influenced orthography used by a French transcriber. Resumptive pronouns are indeed common in many varieties of English, and were most likely re-analysed as subject-referencing pronouns in Melanesian Pidgin (pp. 166; 201–2). However, the English structure that facilitates this re-analysis is not specified. This appears to be the topic construction with anaphoric pronominalization (Givo´n 1976; p. 185), for example The man, he came. The structures of the type I took him and They give him taro (p. 167) are said to have served as a template for the re-analysis of him as a transitive suffix in Melanesian Pidgin. However, there may be two other, more likely sources. One is the afterthought topic shift (Givo´n 1976; p. 157), for example The man, I saw him ! I saw him, the man. Another one is English Foreigner Talk (Ferguson 1975; Mu¨hlha¨usler 1997; pp. 97–8), for example No see um man you say/Me no look him man you say (‘I have not seen the man you are talking about). These two sources might even account for the occurrence of a transitive suffix derived from him in other English-lexified pidgins and creoles for which transfer from their substrate languages cannot be invoked (Avram 2004; pp. 97–8). When discussing substrate reinforcement in Hawai’i Creole, Siegel states that the Portuguese possessive construction with de/do ‘is not used with pronominal possessors’ (p. 198). In fact, de can be used with second and third person pronominal possessors: for example a casa de voceˆ / voceˆs (‘your sg. polite / pl.

REVIEWS 319

Reviewed by Andrei A. Avram University of Bucharest, Romania E-mail: [email protected] doi:10.1093/applin/amq008 Advance Access published on 19 March 2010

REFERENCES Avram, A. A. 2004. ‘Atlantic, Pacific or worldwide? Issues in assessing the status of creole features,’ English World-Wide 25: 81–108. Ferguson, C. A. 1975. ‘Toward a characterization of English foreigner talk,’ Anthropological Linguistics 17: 1–14.

Givo´n, T. 1976. ‘Topic, pronoun and grammatical agreement’ in Li C. (ed.): Subject and Topic. Academic Press, pp. 151–88. Mu¨hlha¨usler, P. 1997. Pidgin and Creole Linguistics. Expanded and revised edn. University of Westminster Press.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

house’), a casa dela (< de + ela) / deles (< de + eles) (‘her / their masc. house’). The reduced forms /blo˛/, /blo/ and /bl/ of bilong (‘of’) used by young urban speakers of Tok Pisin (p. 241) should be transcribed as [blo˛], [blo], and [bl] since they represent phonetic realizations. A similar ‘interpretative typo’ occurs on p. 241: /dok blem/, which should appear as [dok blem], a phonetic realization of dok bilong em (‘his dog’), and the same applies to the examples of occasional Australian pronunciations of vowels in contemporary Tok Pisin on p. 245: [aI] not /aI/, [e] for /e/, and [naIm] instead of /nem/ (‘name’). However, the preceding remarks do not detract in the least from the value of the work. To conclude, Siegel has produced an excellent book, authoritative and up-to-date. His scholarship, both in the field of pidgin and creole studies and in that of SLA, is impressive and this is reflected in the extensive scope of this volume. This is an outstanding contribution to our understanding of the emergence and development of pidgins and creoles for which the author should receive ample credit.

Applied Linguistics: 31/2: 320–322 doi:10.1093/applin/amq011

 Oxford University Press 2010

NOTES ON CONTRIBUTORS

Gary Barkhuizen is an Associate Professor in the Department of Applied Language Studies and Linguistics at the University of Auckland, New Zealand. He has worked as an English teacher and teacher educator in the USA, South Africa, and New Zealand. His research interests are in the areas of language teacher education, sociolinguistics, and narrative inquiry. He has published his works in journals such as TESOL Quarterly, International Journal of Bilingualism, System, ELT Journal, and International Journal of the Sociology of Language. He is the author of Analysing Learner Language (2005, Oxford University Press) with Rod Ellis. John Bitchener is Professor of Applied Linguistics at AUT University, Auckland, New Zealand, Co-editor of NZ Studies in Applied Linguistics and President of the Applied Linguistics Association of New Zealand. Research interests include SLA theory, research and practice (especially written and oral corrective feedback), discourse, and genre studies. His work has been published in journals such as Journal of Second Language Writing, Journal of English for Specific Purposes, ELT Journal, Language Teaching Research Journal, Language Awareness, System, Australian Review of Applied Linguistics, and in a number of New Zealand publications. Address for correspondence: School of Languages and Social Sciences, AUT University, Private Bag 92006, Auckland, New Zealand. <[email protected]> Deborah Cameron is a sociolinguist with a particular interest in language and gender. She currently holds the Murdoch Chair of Language and Communication at Oxford University. <[email protected]> Pedro Guijarro-Fuentes is a Reader in Spanish. His research interests are in the interdisciplinary field of Spanish applied linguistics, sociolinguistics, psycholinguistics, and bilingualism. His research projects include a study of copula verbs

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Andrei A. Avram is a Professor of English Linguistics at the University of Bucharest, Romania. He holds a PhD in linguistics from the ‘Iorgu Iordan – Al. Rosetti’ Institute of Linguistics of the Romanian Academy, Bucharest (2000), and a PhD in linguistics from Lancaster University (2004). His research areas are pidgins and creoles, language contacts, and phonology. His publications include On the Syllable Structure of English Pidgins and Creoles (2005, Bucharest, Editura Universita˘tii din Bucures ti), Fonologia limbii japoneze contemporane (2005, Bucharest, Editura Universita˘¸tii din Bucures ti), articles in English World-Wide, English Today, Revue roumaine de linguistique, as well as reviews in Language in Society, Journal of Sociolinguistics, Bibliotheca Orientalis. Address for correspondence: Department of English, University of Bucharest, 7–13 Pitar Mos Str., 010451 Bucharest, Romani.

NOTES ON CONTRIBUTORS

321

and the acquisition of spatial prepositions by L2 learners funded by the British Academy and the Art and Humanities Research Council. He has authored and co-authored numerous articles and book chapters, and his research has appeared in international journals such as Language Learning, Bilingualism: Language and Cognition, Cognition, Studies in Second language Acquisition and International Journal of Bilingualism. Addresses for correspondence: CKY320, Cookworthy Building, Drake Circus, Plymouth, Devon, PL4 8AA, UK.

Ute Knoch is a Research Fellow in the Language Testing Research Centre at the University of Melbourne. Her research interests are in language testing, particularly in the areas of writing assessment, rater training, rating scale development and language assessment for specific purposes as well as language pedagogy and language and identity. She has published in journals such as Language Testing, Language Assessment Quarterly, TESOL Quarterly, and Assessing Writing. Address for correspondence: Research Fellow, Language Testing Research Centre, University of Melbourne, Room 521, Level 5, Arts Centre, Carlton, Victoria, 3052, Australia. Phoenix Lam is a Lecturer in Applied Linguistics at the School of Arts and Social Sciences, The Open University of Hong Kong. Her research interests include pragmatics, corpus linguistics, discourse analysis, discourse particles, spoken discourse, and English language teaching and learning. Her recent publications appear in Discourse Studies and the International Journal of Corpus Linguistics. Address for correspondence: School of Arts and Social Sciences, Room A0722, 30 Good Shepherd Street, Ho Man Tin, Kowloon, Hong Kong. Jason Rothman is Assistant Professor of Hispanic Linguistics and Language acquisition at the University of Iowa. His research is conducted within the generative paradigm and spans several instances of language acquisition including, L1, L2, L3 and heritage language bilingualism. Recent work has appeared in such journals as the International Journal of Bilingualism, Journal of Pragmatics, Language Acquisition, Second Language Research and Studies in Second Language Acquisition. Address for correspondence: 111 Phillips Hall, University of Iowa, Iowa City, IA 52317, USA. <[email protected]> Julia Sallabank is a Research Fellow in Language Support and Revitalisation in the Endangered Languages Academic Programme at the School of Oriental and African Studies, London. She has been conducting language documentation

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Helen Kelly-Holmes is a lecturer in sociolinguistics and new media at the University of Limerick, Ireland. Address for correspondence: 108 Kilteragh, Dooradoyle, Limerick.

322 NOTES ON CONTRIBUTORS

and sociolinguistic research in Guernsey, Channel Islands since 2000, and is currently comparing language policies in small island states. She gained her doctorate at Lancaster University in 2007. She was previously commissioning editor for Applied Linguistics and Language Teaching Methodology at Oxford University Press. Her main research interests are language revitalization, language policy and planning, and orthography development. Address for correspondence: Department of Linguistics, School of Oriental and African Studies, University of London, Thornhaugh Street, Russell Square, London WC1H 0X, UK. <[email protected]>

Alison Sealey is a Senior Lecturer in Modern English Language at the University of Birmingham. Her research interests span discourse analysis, corpus linguistics, and the sociology of language; she was the principal investigator on the ESRC funded research project ‘An investigation into corpus-based learning about language in the primary school’; and she has published extensively in all of these areas. As well as many journal articles and book chapters, she is the author of Childly Language: Children, Language and the Social World, and, with Bob Carter, of Applied Linguistics as Social Science. She is currently completing a textbook for university students on Researching English Language. Address for correspondence: English Department, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. Guoxing Yu is a Lecturer in Applied Linguistics at the University of Bristol. His main research interests are in language testing and assessment, language factors in assessment of subject learning in national examinations, assessment of learning power, and monitoring quality of basic education in relation to students language and literacy development. He has published in Assessment in Education, Educational Research, and two papers in Language Testing. Address for correspondence: Graduate School of Education, University of Bristol, 35 Berkeley Square, Bristol, BS8 1JA, UK.

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Gila A. Schauer is a Lecturer in the Department of Linguistics and English Language at Lancaster University. Her research interests include interlanguage pragmatics, cross-cultural pragmatics, and cross-cultural impoliteness. Her articles have appeared in Intercultural Pragmatics, Language Learning, and System. She is the author of Interlanguage Pragmatic Development: The study abroad context (2009) published by Continuum. Address for correspondence: Department of Linguistics and English Language, Lancaster University, Lancaster, LA1 4YT, UK.

Applied Linguistics: 31/2: 323–324 doi:10.1093/applin/amq010

ß Oxford University Press 2010

READERS DURING 2008/2009

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

The Editors would like to thank the following for undertaking to review contributions submitted to Applied Linguistics. Zsuzsanna Ittzes Abrams, Carolyn Adger, Svenja Adolphs, Katherine Ahrens, Nobuhiko Akamatsu, Reima Sado Al-Jarf, Rasheed Saleem Al-Jarrah, Linda Allen, Harriett A. Allison, Stephen Andrews, Philipp Sebastian Angermeyer, Marta Anton, Sophie Arkoudis, Karin Aronsson, Dwight Atkinson, Elsa Auerbach, Kathleen Bailey, Paul Baker, Wendy Baker, Will Baker, Joe Barcroft, Kathleen Bardovi-Harlig, Francesca Bargiela, David Barton, Ruba Bataineh, Gulbahar Beckett, Nancy Bell, Carol Berkenkotter, Marı´a Asuncio´n Bes, Rakesh Bhatt, Douglas Biber, Laura Black, Adrian Blackledge, Robert Blake, David Block, David Bloome, Susan Bobb, Christine Bongartz, William Bonk, Pietro Boscolo, Melissa Bowles, Shirley Brice Heath, Sarah Briggs, Lindsay Brooks, Kimberly Brooks-Lewis, Catherine Brouwer, Ray Brown, Paul Bruthiaux, Cathy Burnett, Anne Burns, Cade Bushnell, Yuko Butler, Wolfgang Butzkamm, Martin Bygate, Alice Caffarel, Stephen Caldas, Carmen Caldas-Coulthard, Richard Camaron, Lynne Cameron, Suresh Canagarajah, Maria Carreira, Patricia Carrell, Donald Carroll, Ron Carter, Christine Casanave, Holly Cashman, Marcela Cazzoli-Goeta, Asta Cekaite, Jasone Cenoz, Carol Chapelle, Monika Chavez, Katherine Chen, Liying Cheng, Jenny Cheshire, Dorothy Chun, Elaine Chun, Eton Churchill, KimMarie Cole, Jim Coleman, Joseph Collentine, Kathy Conklin, Angela Creese, Jakob Cromdal, Alister Cumming, Mary Jane Curry, Christiane Dalton-Puffer, Fred Davidson, Winifred Davies, Kees de Bot, Peter De Costa, Anna De Fina, Maria C.M. de Guerrero, John H.A.L. de Jong, Robert de Keyser, Ana de Prada Perez, Alice Deignan, Tracey Derwing, Rainer Dietrich, Julie Dockrell, Evelyn Doman, Laura Domı´nguez, Richard Donato, Zoltan Dornyei, Olga Dragoy, Patricia Duff, Kerry Dunne, Fred Eckman, Nick Ellis, Rod Ellis, Dogu Erdener, Elizabeth J. Erling, Fiona Farr, Anwei Feng, John Field, Edward Finegan, Tess Fitzpatrick, Lynne Flowerdew, Keith Folse, Rebecca Foote, Cecilia Ford, Fanny Forsberg, Pauline Foster, Janna Fox, Barabara Freed, Mark R. Freiermuth, Debra Friedman, Pedro A. FuertesOlivera, Jia Gisela, Joseph Gafaranga, Phil Gaines , Blazej D. Galkowski, Colin Gallagher, Adela Ganem, Dee Gardner, Sheena Gardner, James Paul Gee, Alexandra Georgakopoulou, Ghazi M. Ghaith, Andrew Goatley, Manuela Gonza´lez-Bueno, Hugh Gosden, Barbara Graves, Mike Grenfell, Janet Grijzenhout, Laura Gurak, Damodar Gurrapu, Atiqa Hachimi, Kira Hall, Megumi Hamada, Mary Hamilton, Liz Hamp-Lyons, Linda Harklau, Michael Harrington, Kevin Harvey, David Hayes, Carl Haywood, Lianzhen He, John Hellerman, Christina Higgins, Pamela Hobbs, Adrian Holliday, Janet Holmes, Sue Hood, Yuri Hosoda, Juliane House, Rebecca Howard, Tsung Hua Linda, Zhu Hua, Thom Hudson, Jan Hulstijn, Elizabeth Hume,

324 READERS DURING 2008/2009

Downloaded from applij.oxfordjournals.org by guest on December 31, 2010

Susan Hunston, Rick Iedema, Ofra Inbar, Akihiro Ito, Suzsanna Ittzes Abrams, Mark Jary, Malgorzata Jedynak, Ulrike Jessner-Schmid, Joan Swann, David Cassels Johnson, Karen Johnson, Zohar Kampf, Shumin Kang, Su-Ja Kang, Yasuko Kanno, Gabriele Kasper, Greg Keating, Susan Kemper, Kerrie Kephart, Richard Kern, Yanghee Kim, Celeste Kinginger, Keiko Koda , Tom Koole, Irene Koshik, Nelya Koteyko, Michele Koven, Claire Kramsch, Ryuko Kubota, Kon Kuiper, Kumaravadivelu, Wendy Lam, Yuen Kwan Wendy LAM, Martin Lamb, Zak Lancaster, Diane Larsen-Freeman, Cynthia Lee, Sang-Ki Lee, P.P.M. Leseman, Constant Leung, John Levis, Danli LI, Defeng Li, Tony Liddicoat , Jason Miin-Hwa Lim, Chien-Jer Charles Lin, Seth Lindstromberg, Jianda Liu, Maria Pilar Agustin Llach, Gillian Lord, John Paul Loucky, Graham Low, Wander Lowie, June Luchjenbroers, Andrew Lynch, Ernesto Macaro, Ronald Macaulay, David Machin, Peter MacIntyre, Sally Magnan, Rosa Manchon, David Marsh, Ron Martinez, Paul Matsuda, Janet Maybin, Steve McCafferty, Michael McCarthy, James McCroskey, Kim McDonough, William H McKellin, Tim McNamara, Paul Meara, A J Meier, Julia Menard-Warwick , Susie Miles, Liz Miller, Philip Monahan, Silvina Montrul, Rosamund Moon, Pat Moore, Junko Mori , Janus Mortensen, Kristian Mortensen, Scott Myers, Florence Myles, Terry Nadasdi, Shigeko Nariyama, Paul Nation, Hilary Nesi, Matt Newman, Jonathan Newton, Hanh Nguyen, Christiane Nord, Catrin Norrby, Sarah North, Bonny Norton, David Nunan, Kay O’Halloran, Anne O’Keefe, David Oakey, Seth Ofori, William O’Grady, Amy Ohta, Anne O’Keefe, David Olsher, Lourdes Ortega, Mits Ota, Maryann Overstreet, Norbert Pachler, Iliana Petrova Panova, Myung-Kwan Park, Anne Pomerantz, Suzanne Marie Prior, Patrick Proctor, Victoria Purcell-Gates, David Qian, Vaidehi Ramanathan, Ben Rampton, Anne Rasanen, Dorit Ravid, John Read, Gisela Redeker, Vera Regan, Katherine Rehner, Hayo Reinders, Brian Richards, Keith Richards, Shelagh Rixon, Julio Roca de Larios, Karen Roehr, Carsten Roever, Stephen Ross, Jason Rothman, Suzanne Rott, Elizabeth Rowley-Jolivet, Christopher Ruehlemann, Otto Santa Anna, Shannon Sauro, Rachel Schiff, Barbara Schutz, Mike Scott, Alison Sealey, Paul Seedhouse, Mike Sharwood-Smith, Saad shawer, Younghee Sheen, Ling Shi, Meryl Siegal, James Simpson, Peter Skehan, Stef Slembrouck, Bernard Smith, Catherine Snow, Jeong-Bae Son, Nina Spada, Robyn Spence-Brown, Bernard Spolsky, Neomy Storch, Michael Stubbs, Gretchen Sunderman, John Swales, Michael Swan, Joan Swann, Naoko Taguchi, Christine Tardy, Celia Thompson, Geoff Thompson, Paul Thompson, Joshua Thoms, Steve Thorne, Cristian Tileaga, Brian Tomlinson, Jacqueline Toribio, John Truscott, Christoph Unger, Erma Ushioda, Theo van Leeuwen, Margaret van Naerssen, Eija Ventola, Carla Vergaro, Marjolijn Verspoor, Caroline Vickers, Daniel Villa, Johannes Wagner, Steve Walsh, Rob Waring, Mark Warschauer, Stuart Webb, Sarah Weigle, Regina Weinert, Eddie Williams, Cynthia Wiseman, Kate Wolfe-Quintero, David Wood, Devon Woods, Alison Wray, Junko Yamada, Foong Ha Yap, Richard Young, Verhawn Ashanti Young.