Language Studies: Stretching the Boundaries [Unabridged] 1443839728, 9781443839723

Citation preview

Language Studies

Language Studies: Stretching the Boundaries

Edited by

Andrew Littlejohn and Sandhya Rao Mehta

Language Studies: Stretching the Boundaries, Edited by Andrew Littlejohn and Sandhya Rao Mehta This book first published 2012 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2012 by Andrew Littlejohn and Sandhya Rao Mehta and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-3972-8, ISBN (13): 978-1-4438-3972-3

TABLE OF CONTENTS

List of Pictures........................................................................................... vii List of Figures .......................................................................................... viii List of Tables.............................................................................................. ix Introduction ................................................................................................. 1 Andrew Littlejohn Section I: Concepts Considered Chapter One............................................................................................... 10 Who is Stretching Whose Boundaries? English Language Studies in the New Millennium Sandhya Rao Mehta Chapter Two .............................................................................................. 26 Language and Group Identity: Some Social Psychological Considerations Itesh Sachdev Chapter Three ............................................................................................ 43 Procedures for Translating Culturally Specific Items James Dickins Chapter Four.............................................................................................. 61 Proverb Translation: Fluency or Hegemony? An Argument for Semantic Translation Abdul Gabbar Al-Sharafi Chapter Five .............................................................................................. 75 Dialogue Systems: Stretching the Boundaries of Pragmatics and Discourse Analysis Radhika Mamidi

vi

Table of Contents

Chapter Six ................................................................................................ 93 The Role of Forensic Linguistics in Crime Investigation Anna Danielewicz-Betz Chapter Seven.......................................................................................... 109 University English Studies in Multilingual Contexts: What are the Prospects? James A. Moody Section II: Languages Considered Chapter Eight........................................................................................... 126 How English Grammar has been Changing Geoffrey Leech Chapter Nine............................................................................................ 147 Digging for New Meanings: Uncovering a Postcolonial Beowulf Jonathan Wilcox Chapter Ten ............................................................................................. 162 "These words are not mine. No, nor mine now." Poetic Language Relocated Sixta Quassdorf Chapter Eleven ........................................................................................ 177 Stretching the Boundaries of English: Translation and Degrees of Incorporation of Anglicisms Paola Gaudio Chapter Twelve ....................................................................................... 190 The Arab Body Metaphor in Contemporary Arabic Discourse: An Exploratory Study Abdullah al Harrasi Chapter Thirteen...................................................................................... 208 Students as Authors: Textual Intervention in Children's Literature Rosalind Buckton-Tucker Contributors............................................................................................. 218 Index........................................................................................................ 221

LIST OF PICTURES

CHAPTER FIVE Picture 5.1 Picture 5.2 Picture 5.3 Picture 5.4

Gold sword hilt guard with zoomorphic decoration Gold helmet cheek guard with zoomorphic decoration Silver-gilt animal terminal, probably from helmet Inscribed gold strip

LIST OF FIGURES

Figure 3.1 Figure 4.1 Figure 9.1 Figure 11.1 Figure 11.2 Figure 11.3

The Brown Family: the Brown, Frown, LOB and FLOB Corpora Procedures for translating culturally specific items Semiotic representation of the three concepts of sign, proverb and translation Architecture of a simple dialogue system Interactive model of language processing Dialogue modelling frameworks

LIST OF TABLES

CHAPTER THREE Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 3.7

Decline in the use of modal auxiliaries in written English Increase in the use of semi-modals in written English Increase in the use of the present progressive in written English Declining frequency of the passive breakdown by subcorpus Comparison of SEU-mini-sp and ICE-GB-mini-sp: modals in spoken BrE (Provisional figures) Comparison of SEU-mini and ICE-GB-mini: some ‘semi-modals’ in spoken BrE The changing semantics of may, based on an analysis of 1 in 3 examples found (provisional): American and British English CHAPTER SIX

Table 6.1

Frequencies in HyperHamlet, BNC, COCA and COHA

INTRODUCTION ANDREW LITTLEJOHN

As a defining characteristic of what it means to be human, the use of language plays a central role in virtually every form of human activity. From the moment when our first cries, as a newly born child, begin to mimic the melody of the speech around us, we begin an involuntary process in which language progressively ties us into our immediate community and into the wider society. Linked as we are in this way, language functions as a cornerstone in the construction of our identity amongst others and in the relationships we build. It takes a central role in facilitating every enterprise we undertake, alone or together with others, as our language enables the formulation and communication of ideas. It permits us to reflect on our past, analyse our present and plan our future, creating the central thread through our own biographies as we endeavour to make sense of the narratives of our lives. And, beyond the confines of our own personal possibilities, language enables us to take part in the transmission of our own culture, its maintenance, evolution and perhaps eventual demise in the face of rival formulations of ideas. Given its pivotal role in virtually every action we take, it is perhaps surprising, then, that we could even consider that the study of language would have any boundaries. The notion of stretching the boundaries, the subtitle of this collection of papers, may thus sound a somewhat odd idea. Yet, the study of language has traditionally been focussed on language itself, its analysis and description, and those broader links which have been formed have most frequently been found as extensions of other disciplines – the psychology of language, for example, is often considered an extension of psychology; the sociology of language, an extension of sociology; and so on. In this sense, then, language studies, as an academic area of interest, may, historically, have imposed boundaries upon itself, perhaps in trying to carve out a discernible space as it competed for academic recognition and status as a science—linguistic science—largely leaving the interrelationship of language with other areas of human knowledge for others to elucidate.

2

Introduction

Such, however, is not the nature of modern-day language studies. The field is now characterised by an intense cross-fertilisation of ideas, often from distant disciplines. Thus, an area which was once typified by a consensus over its domain of reference is now characterised by numerous, often competing frames, and a fragmentation of what is considered its research focus. Examples abound: the development of computers has brought us ‘computational linguistics’; the study of politics and economics has shaped an emerging field of ‘critical discourse studies’; theories of argumentation have been brought into theories of translation to establish an emerging field of ‘translation quality assessment’; the formal study of language has established links with brain research and with sociology to bring us recent concepts such as ‘gendered speech’; to name but a few developments. The net effect of these cross-fertilisations is that there is now considerable discussion over the parameters of language-related disciplines, particularly literature, linguistics and translation, and a profusion of work at what would once have been seen as the ‘margins’. The ‘centre’ or rather ‘centres’ of interest are thus now being redefined. It was the desire to explore some of these developments that inspired a conference, hosted at Sultan Qaboos University, Oman, in 2010. Each of the papers included in the present collection had its origins in this conference, although they have since undergone extensive debate and refinement. Together, the papers present a rich palette of themes, signalling some of the diverse work which, today, can be said to comprise the discipline of language studies. Two major points of focus divide the collection into its two sections. The first, entitled Concepts Considered, reviews some of the theoretical concepts which underpin different aspects of language study, while the second section, entitled Languages Considered, pursues the application of theoretical concepts in the context of a specific language in use. In the opening paper of the first section, Sandhya Rao Mehta takes a broad view on the evolution of language studies. Although Mehta’s main concern is with the development of English language studies, all of the themes she identifies will resonate with researchers in any language. In her paper, Mehta shows how the ‘meta-narrative’ in English studies has shifted from the concern I mentioned earlier, issues of language itself, towards something much more diffuse and dynamic: users of language, often in geographically diverse parts of the world, who assert a variety of Englishes, for their own purposes in an ever expanding number of contexts and modes. A major aspect of this is the emergence of ‘e-English’, which poses new challenges for language studies as we grapple with the analysis of virtual identities and virtual language—dynamically changing users

Language Studies: Stretching the Boundaries

3

who equally dynamically shape new genres of language use. Mehta also points to the development of Englishes in literature, where the traditional focus on British and American literature is being forced to give way to the recognition of an expanding range of writers, from various parts of the world, now writing in English. She highlights the impact of multiple migrations, and the difficulties this presents in trying to establish precisely who is writing for which community and in which literary tradition, as transient diasporic communities lead “identities and languages [to] coalesce into increasingly complex layers” (p12). The relationship between language use and identity is also the concern of Sachdev, who shows how language simultaneously reflects and creates group identity. Data from street interactions in bilingual Tunisia, for example, showed how pedestrians’ choice of language, when replying to requests for directions from other Tunisians, depended not only on the language of the request (Arabic or French) but also on the ethnic origin of the requester (‘brown’, ‘white’ or ‘black’). Code-switching in Arabic/ French, Sachdev found, was evident with ‘brown’ Tunisians but never occurred with ‘white’ or ‘black’ compatriots, regardless of the language of the request. Sachdev argues that code-switching in conversation with ‘brown’ Tunisians thus simultaneously denoted both status (i.e. a higher level of education) and group solidarity. As Sachdev says, “linguistic choices were identity choices…permeating even the briefest interactions” (p30). This, and additional data from Canada, Bolivia, India and the Indian and Pakistani diasporas, show the strength of the link between identity and language use, a finding which, Sachdev argues, underscores the importance of supportive policies to maintain the languages of minorities. The complexity of communication across languages and cultures is the concern of the next two authors, both of whom debate issues surrounding policies in translation. One of the key difficulties facing any translator is how ‘cultural items’ should be handled, in particular, how far they should be rendered ‘domestic’, such that any trace of the source culture is removed, or how far they should maintain their ‘foreign’ character— comprehensible but still noticeably ‘foreign’. Dickins provides a wide ranging analysis of the approaches advocated by various theorists when encountering such items, and sets out a conceptual framework to show how these approaches stand in relation to each other. The resulting grid should provide significant support to translators who need to clarify for themselves their choices in translation. In the next paper, however, AlSharafi advocates a particular view in translation policy, arguing for a ‘semantic’ (literal) translation of items such as proverbs, rather than a ‘communicative’ translation which abandons any trace of the source

4

Introduction

culture and text in favour of assimilation into the target language and culture. At stake here, for Al-Sharafi, is not simply a matter of translation efficacy. Our attitudes towards the translation of cultural items can betray a hegemonic and ethnocentric attitude towards other cultures “because if we translate in ways natural to readers, then the element of learning from other cultures will disappear” (p68). The need to establish a framework for handling language is also a concern for the next two papers, although in quite different areas, reflecting the breadth that the field of language studies now encompasses. Mamidi presents a stimulating analysis of the problems surrounding the design of dialogue systems—that is, systems which enable humans to interact with computers using natural language. Mamidi shows how the subtleties and complexities of human speech, particularly with respect to pragmatics and language in discourse, present formidable challenges for systems design. At present, only limited success has been achieved in some domains with only a restricted set of interactional outcomes possible, but Mamidi successfully shows how our knowledge of the workings of language, both as form and as meaning, will be vital in building more complex systems capable of handling interactions with humans in an ever more flexible manner. A very different challenge concerns Danielewicz-Betz, however, as she once again returns to the earlier themes of language and identity—this time in the context of crime investigations and the use of forensic linguistics to identify the perpetrator or victim. Like Mamidi, Danielewicz-Betz finds only limited success to date in the application of linguistic knowledge, and doubts whether we will ever be able to establish the existence of a ‘linguistic fingerprint’ as reliable as DNA evidence. Despite these limitations, Danielewicz-Betz shows that linguistics can play a significant role in eliminating individuals from an investigation and in ensuring accurate statements of witness evidence. In the final paper of this section, Moody returns to many of the issues raised by earlier writers concerning language and identity in multilingual communities, but in the context of the development of programmes of study in university English departments. Much contemporary discussion in English language teaching, argues Moody, evinces a “diffusion-ofEnglish” model which emphasises native-speaker standards in language use, a one-size-fits-all approach to teaching materials, and a failure to recognise the complexity of the need, perception and use of English in multilingual communities. University English departments, Moody suggests, are frequently characterised by a “fragmentation of courses, an emphasis on theory and knowledge over practice, misconceptions about

Language Studies: Stretching the Boundaries

5

students’ motivation and assumptions of their inadequacy” (p114), all of which positions the student as an outsider and a failure in their own society. In a strongly argued case, Moody sets out a proposal for the development of a degree programme in English studies which would more fully take account of an ecological view of English—as it is seen, used and needed in a specific community. In the second section of the collection, Languages Considered, a number of writers discuss many of the concepts raised earlier, but in the context of research into specific languages. The section opens with a paper by Geoffrey Leech which chronicles the changing nature of English grammar. Noting that many native speakers of English believe that standards of grammar use are “deteriorating”, Leech argues that it is more helpful to see changes in grammar as adaptations to new influences and new possibilities in communication. Drawing on extensive data taken from corpora representing a full generation of language users—approximately 30 years—Leech presents intriguing findings which show how, for example, modals such as may, must, need(n’t), ought to, and shall, are becoming much less common, while forms such as be going to, need to and have to are increasing in use. Leech argues that there are three forces at work here: grammaticalisation (in which lexical phenomena evolve into grammatical phenomena), colloquialisation (in which spoken habits infiltrate written forms) and Americanization (in which American English grammatical constructions become more common in British English). Before language teachers rush to redesign their syllabi, however, Leech suggests caution: these changes are happening very slowly, although attention should always be paid to frequency of use. In this, corpora offer vital support to language teaching professionals, he argues. Jonathan Wilcox also provides a historical perspective on language, but to a much more distant point—that of the Anglo-Saxon tale of Beowulf. Wilcox’s paper presents a fascinating account of just how far language studies can be stretched, as he unites the three fields of literature, language and archaeology to ‘dig for new meanings’ in the poem. Examining Seamus Heaney’s controversial translation of the poem, which incorporates language references to Catholic Ulster, Wilcox shows how choices in translation prove central to a positioning of the poem within contemporary postcolonial concerns and contested claims to the past. Just how contested the past can be was clearly brought to the surface by the discovery of the Staffordshire Hoard—a treasure trove of Anglo Saxon booty found in a field in 2009. In his account of local reactions to the find, Wilcox shows how a fictional history was woven around the objects, to create the image of the region’s glorious past. For Wilcox, this was largely

6

Introduction

prompted by “present-day Britain’s awkward status within a postcolonial world” in which it “lacks a clear sense of independent and unifying nationhood” (p160). In interpreting these reactions, however, Wilcox shows how the language of Beowulf and Staffordshire Hoard illuminate each other, allowing a closer reading of the significance of both the poem and the hoard, and suggesting that fragments of an Anglo-Saxon past— whether linguistic or material—can be deeply revealing of contemporary desires. In the next paper, Quassdorf also shows how the literary language of the past may be used for contemporary purposes, as the original writer’s intention is recast to meet the modern-day user’s purpose. Quassdorf takes the reader through a new database, HyperHamlet, which allows crossreferencing from Shakespeare’s play to samples of modern language, to show how people draw on quotations, and relocate the sentiment expressed. She shows how frequently quoted lines from the text may come to deviate from the contextual, pragmatic and formal qualities of the original, thus achieving a growing independence from the source and a conventionalisation of use. She suggests users “relocate, re-apply and even misuse his words” such that they become part of the langue. In this way, we can see how Shakespeare has contributed to much to the shaping of English, as has so often been claimed. Gaudio is also concerned with notions of incorporation, but this time from English into another language—Italian—in the form of anglicisms. The challenge for Gaudio is to find a basis for the translator’s decisionmaking in the face of anglicisms in the source text—should they be translated, glossed or left as they are? Rejecting an ‘ethical’ approach which considers issues of a ‘domesticating’ or ‘foreignising’ translation (cf Dickins and Al-Sharafi, this volume), Gaudio argues that the correct basis for translation is to look at what users actually do with the language. Anglicisms need to be analysed to determine if they are unincorporated, semi-incorporated or fully-incorporated into the target language, and an objective decision can then be made on the need for translation or glossing. To support this analysis, she argues for the cross referencing of parallel corpora from the source and target languages. Drawing on the multilingual versions of the Official Journal of the European Union, Gaudio gives a number of examples of anglicisms in Italian and shows how an informed decision about translation can now be made. Al Harrasi makes extensive use of corpora in his paper to analyse how one particular metaphor—the Arab world as a human body—is used in online discussions. His analysis is set within a conceptual theory of metaphor, which presents metaphor as a device through which we map one

Language Studies: Stretching the Boundaries

7

domain of experience, usually a concrete experience, on to another domain, normally an abstract one. Citing numerous examples, Al Harrasi shows how political discourse is peppered with references to human organs, bodily ailments, and medical procedures as a way to analyse the political difficulties which the Arab world faces. The point which Al Harrasi stresses, however, is the significance of this linguistic ideological framing and the implications it has in directing any action which is undertaken. The body politic metaphor, he argues, “leads to the dismissal of real political phenomena” and an emphasis on a fictional homogeneity which can “marginalize or even eliminate political entities that are different from the majority” (p202). Clearly, such metaphors work in the interests of some and against the interests of others. In the final paper in the collection, Buckton-Tucker brings us back to the issues surrounding the teaching of language, echoing one of the main themes of the collection: that language rightfully belongs to its users. Drawing on an innovative technique known as ‘textual intervention’, she gives us many examples of how students can be engaged in taking ownership of their reading and their writing. Through the use of the technique, which involves students in rewriting or extending literary texts according to their own interpretation or cultural context, Buckton-Tucker shows how students may simultaneously develop a better understanding of a literary genre, improve their language skills, and find their own voice as creative language users. Taken together, the papers in this collection provide an absorbing, rich array of subjects touched by the centrality of language. Encompassing themes from the study of social psychology, translation theory, computer science, forensics, educational policy, language change, archaeology, and literature, to name but a few, the collection shows that a concern with the role of language continues to expand through cross-fertilisation with a limitless number of other disciplines. For the field of language studies, this has meant that issues which once would have been considered marginal have now become central, as the boundaries of relevance continue to be stretched.

SECTION I CONCEPTS CONSIDERED

CHAPTER ONE WHO IS STRETCHING WHOSE BOUNDARIES? ENGLISH LANGUAGE STUDIES IN THE NEW MILLENNIUM SANDHYA RAO MEHTA

Abstract Exploring one of the most significant features of language studies at the turn of the twenty-first century, this paper examines the way in which the centres of English speaking communities have expanded to include a variety of participants in geographically diverse parts of the world. It also points to the increased focus on the use of language in the virtual world, making possible the creation of parallel communities online, as well as other technologically driven forms such as texting and instant messaging. This study traces the way in which the focus of research has shifted from the nature of language itself to the changing nature of the English speaking world and the impact it has had on the language as well as its literatures. The implications of these emerging findings on related areas of ELT and classroom methodologies are then discussed to identify ways in which language teaching could be linked to rapidly changing research.

Keywords: English language studies, speech communities, English and the internet, blogging, texting, postcolonial English, circles of English

Introduction In an iconic scene in the famous children’s tale Alice in Wonderland, Alice wanders into the Mad Hatter’s tea party where it is suggested that they ask riddles to pass the time. The first riddle “Why is a raven like a writing desk?” starts off an intense conversation that points to the

English Language Studies in the New Millennium

11

complexity of language formulations, meanings and representations in the real world: “I’m glad they’ve begun asking riddles–I believe I can guess that,” She added aloud. “Do you mean that you think you can find out the answer to it?” said the March Hare. “Exactly so,” said Alice “Then you should say what you mean,” The March Hare went on. “I do,” Alice hastily replied; “at least–at least I mean what I say–that’s the same thing, you know.” “Not the same thing!" said the Hatter. “You might just as well say that “I see what I eat” is the same thing as “I eat what I see!” (Carroll, 1865, p. 81)

Long viewed as an example of linguistic and philosophic cleverness, this conversation continues to highlight the dynamics of language and the relations between word and meaning and the disorder behind the seemingly ordered world of Victorian society. More significantly, at a linguistic level, it also points to the dialectic relationship between word and meaning, 'language and reference’ (Flescher, 1969) that continued to dominate philosophical and linguistic studies at this time with their focus on the relation between syntax and semantics as well as sign and signifier. While the repartee takes place in the Freudian world of Alice, or, Carroll’s subconscious, it is indicative of the dynamic world of word play that also characterizes the way in which modernism came to be involved with studies in language use.

Modernism and Language Studies Modernism’s preoccupation with form and meaning and the changing phenomenological emphasis on the word have been constantly seen as the central terms of reference in language studies in the twentieth century. Influenced, in turns, by scientific studies of the human mind (leading to cognitive theories of language), of psychology (leading to psycholinguistics) and studies in emerging sociology, language studies came to be defined by the way in which it could be learnt, used and taught as an unchanging commodity whose intrinsic value remained fundamentally unchanged. This was largely coupled with research based on Harold Palmer’s ‘methods teaching’ (1924), Skinner’s behaviourist studies (1957), Piaget’s developmental psychology (1971), Vygotsky’s social constructivist theories (1962) and reworkings of power in language use as construed by Fairclough (1989) among many others. While the central conflict in

12

Chapter One

language emerged as the struggle between the Chomskyian notions of ‘universal grammar’ (1965), Pinker’s idea of latent language (1984) and the more essentialist, Arnoldian approach of language being itself beautiful and essentially ennobling, the predominant approach remained, to a large extent, the division between language and speaker, text and reader, even commodity and consumer. In the last two decades of the twentieth century, these links between words and meanings began to assume more complex layers of possibilities, coinciding with a variety of changes in the political, social and cultural spheres. If a single reason could account for the shift from the hitherto defined areas of communication to language as empowerment, the gradual way in which the globe was becoming ‘flat’ (Friedman, 2005) in many senses of the word, could be a significant one. The shift from language as something that could be simply learnt, to one that could change, be used, re-used and in fact, even re-created, can perhaps be seen as the most significant movement in the years leading to the new millennium. Forces such as the rise of postcolonial societies, increasing movements of people and goods, expanding communication channels and of course, the software revolution can be seen to re-define the way in which language studies have emerged as empowering the user, the endconsumer, thus redefining its usage, very often pushing the boundaries of language itself to create new movements and new worlds (such as the virtual one). Nowhere is this shift more apparent than in the way in which English itself has been re-defined to align it with the changing world.

English in a ‘Flat’ World In very clear terms, the metanarratives of English language studies have shifted focus from being centred around well established issues of speech, grammar and vocabulary to one that is dynamic, shifting and open to a variety of emerging trends, however fleeting or permanent they may be. In this context, perhaps, an important question which arises is, who exactly defines the boundaries in language studies and who stretches them? However debatable this issue may be, it is possibly inarguable that the major onus of changing the ways in which English is being viewed in the contemporary world rests primarily on the individuals who are actually using it, both across geographical as well as virtual spaces. That the English language has spread far wider than ever before in its thousand year history has been established by a variety of studies including those of David Crystal and David Graddol. An important reason for the shift in focus from viewing English as a static whole to varieties of “Global

English Language Studies in the New Millennium

13

English” (Crystal, 1980) is the perceptible shift in the character of the language itself. Graddol, in a study conducted in 1997, established that, while native English speakers could, in fact, reduce in absolute numbers given the rapid fall of population among them and the corresponding rise in the population of Mandarin and Hindi speakers (Graddol, 1997, p. 5), it is the perceptible shift of English being used as a second, third and even foreign language that has established it as the most widely spoken language in the world at the turn of the new millennium. Graddol comments that …in many parts of the world today, as English is taken into the fabric of social life, it acquires a momentum and vitality of its own, developing in ways which reflect local culture and languages, while diverging increasingly from the kind of English spoken in Britain or North America. (Graddol, 1997, p. 4)

Peter Strevens takes up Crystal’s concept of the number of English speakers by stating that “…the figures tell us that while English is used by more speakers than any other language on Earth, its mother tongue speakers make up only a quarter or a fifth of the total” (Strevens, 1992, p. 28). The sheer variety of speakers has been well documented by a number of studies in various parts of the world, initiated by Braj Kachru (1982) and taken up by Cecil Nelson (1996), Peter Strevens (1992) and Yamuna Kachru (2008) among many others. Braj Kachru’s interpretation of this widening of the centres of English takes the form of the expanding circles in which speakers belong variously to the first, second and emerging circles (Kachru, B. 1982, p. 356). This division is re-interpreted by Tripathi (1998) by removing the implicit codes of race and region from these circles and Modiano, whose focus shifts to the way in which English speakers from outside the centre communicate with each other. Jennifer Jenkins shows how Modiano re-formulates English speakers as those inhabiting a centre where the speakers are able to communicate proficiently in English as an international language with the second circle consisting of those who speak English as a first or second language. The third group includes those who are learning English and the fourth, those who know no English at all (Jenkins, 2003, p. 22). In this model, the movement, of course, is inwards, towards language proficiency. This substantial shift in language studies from its historical focus on its own forms and meanings (or lack of them), to the ever widening possibilities of language use that has defined the first decade of the twenty-first century is a shift that may be indicative of the way in which the agenda for future research may also move. The implications of such

14

Chapter One

shifts are equally profound, for, with the shift in agency from the language itself in its multiple manifestations to its users in their various identities, the paradigms of English studies continue to be re-defined. At this social level, the dialectics of language studies are most clearly indicated in the shift from the common coinage of ‘peripheral’, ‘marginal’ and/or ‘other’ (Kachru, 1982) to the more flattened and less geographically explicit concepts such as 'pluricentricism' and the very simple ‘Englishes’ or the all-pervasive concept of ‘Global English’. In this way, the trend towards the empowerment of agency appears to reach its logical purpose, for, if there is one underlying feature that can embrace the various currents of English as it continues to be used across the world in a variety of ways and for various purposes in the real and virtual world, it is the focus on the individuals who have managed to push themselves from being perceived as marginal to becoming identified as, at least, one of the centres. Although it may be an exaggeration to say that English is being re-created, it may not be very far from the truth to assume that people affecting these changes come from, literally, different parts of the worlds, and in fact, increasingly belong to another world altogether–the virtual one.

The Emergence of e-English Research in various areas of language application in the last few decades has implicitly recognized the power of language users to expand the boundaries of their own creations. This is most in evidence in the creative uses of English in the media as well as in the virtual world of the internet. Aptly referred to as the “post-literate culture” (de Castell, 1996, p. 398), this focus on the way in which technologies have redefined communicative forms for universal consumption has been a part of many current debates on language use. English, again, remains at the centre of this debate as it is the language in which most software was created and the language in which commercial activities primarily take place. As recently as December 2011, the number of people using the internet in English was estimated to be around 537 million as opposed to its closest competitor, Chinese, whose users totalled 444 million (Discovery news, Dec. 29, 2011, np.). The internet, using as it does, various different forms of language, mostly informal, redefines the way in which language could be used as a mode of expression when used in the context of blogging, chatting and texting. While the use of technology in the process of language learning was a common strategy in the middle of the twentieth century with language laboratories relying on tape recorders and microphones, the internet added

English Language Studies in the New Millennium

15

a whole new dimension to this encounter by becoming more interactive, bringing the participant to the centre of the language process (Singhal, 1997). The internet allowed for a sudden transformation of the ‘learners’ enabling them to create communities and multiple identities, further enabled by the anonymity offered by the wireless medium. In a society where individuals negotiate effortlessly between the virtual and real worlds, the communities themselves are redefined as they cross the geographical boundaries of states and nations, crossing over even the borders of social and political hierarchies. Media studies such as those of Durant and Lambrou (2009) trace the notion of ‘secondary orality’ as being manifested in the early twentieth century media of television, radio and telephone but then point to the subsequent rise of the internet as an important stage in the development of communication platforms. They suggest that [t]he internet revolution has also moved media language towards multimodal kinds of textual organization. In multimodal discourse, images, written text, music and sound combine and function together. But these changes are as yet incomplete. We don’t yet know how they will develop or turn out. (Durant and Lambrou, 2009, p. 5)

The way in which television language, and more recently the registers on the internet vary has been extensively worked on in terms of the terminology used (Graddol, Cheshire and Swan, 1994; Swales, 1990) as well as through studies on specific uses of language, in particular media related situations (Van Dijk, 1985). This coalescing of genres made possible by the internet feeds into animated forms of discourse that may never be able to be labelled using a specific register. Best exemplified by blogging, the internet has become a significant example of the way in which language continues to be stretched across vocabulary choice, narrative techniques and sentence structures (Rettberg, 2008). Studies on blogging examine the way in which personal narratives and responses emerge as particular voices reacting, not only to a given article, but also to each other as the correspondence progresses. Humour, it is believed, very often becomes part of the thread of reactions to blogs, as also insults and personal attacks in the blogging world (Durant and Lambrou, 2009, p. 135). Research in the various areas of language applications in the virtual sphere have indicated the way in which expressions change as they move from face-to-face discourse to the more anonymous ones conducted online (Crystal, 2001; Baron, 2008; Wright and Street, 2005).

16

Chapter One

While much of the conversation in the virtual world remains between individuals who choose to stay behind changing identities, the more creative possibilities of such a conversation remain largely unexplored. Commonly referred to as ‘netspeak’, the language of the internet is increasingly seen to be an attempt by particular communities to establish specific identities that may often be inaccessible to the mainstream world. Studies in this area include one that looks into the way young girls create closed groups using “interactive writing and the exchange of additional digital information, such as image files and web addresses…enabling these young people to develop sophisticated and marketable skills” (Merchant, 2001). Another one explores the issue of teenagers’ negotiation of identities through language. Using a variety of sources such as Robin Lakoff’s and Deborah Tannen's work on gender and language, a conclusion regarding strategies of language is arrived at: A study of 2692 messages of internet discussion groups finds that groups dominated by females tend to ‘self-disclose’ and avoid or attempt to reduce tension (Savicki, 1996). Similarly, Herring (2000) finds that women are ‘more likely to thank, appreciate and apologise, and to be upset by violations of politeness’ (Herring, 2000). In contrast, discussion groups dominated by males tend to use impersonal fact oriented language (Savicki, 1996) and males seem less concerned with politeness and sometimes violate expected online conduct. (Huffaker and Calvert, 2005, np)

Whatever the quality of writing and responding in the blogging world, it doubtless inspires tremendous creativity and literary inventiveness. By being able to go beyond the limitations of customary rhetoric, online communities have been able to create ways of expression that are novel in their ideas, sharp in their immediacy and focused in their content. Whether this will remain an established form of communication and whether the English of the anonymous online community will integrate itself into accepted language codes may remain controversial and temporarily unsolvable. Crystal’s characteristic response to these emerging trends could perhaps point to a possible way of looking at these developments. In an interview with the BBC in 2010, he reacts to the changing use of English on the internet by saying, “[t]he internet is an amazing medium for languages. Language itself changes slowly but the internet has speeded up the process of those changes so you notice them more quickly” (Crystal, 2010).

English Language Studies in the New Millennium

17

Texting in English Nowhere is this speed in communication strategies more apparent than in the way in which technology merges with the traditional form of literature to create completely new versions of established literary forms. Texting, unlike the broader possibilities of the internet, is known only for its brevity and resulting ‘corruption’ of grammar and spellings. Studies including those of Brown-Owens (2003) and Lenhart et al. (2008) have dwelt extensively on the effects of texting on the levels of English accuracy in direct response to Lee (2002) who condemns texting as an example of the “continuing assault of technology on formal written English” (Vosloo, 2009, np). Vosloo’s study on the effects of texting on school children explores the way in which the creative use of language could lead to a wider sense of creativity and educational possibilities, forcing educationists and the community to broaden the definitions of what it means to be literate. By incorporating studies which show that children have an increased sense of the way in which language functions and understand the connections between sound, word, meaning and text, Vosloo attempts a critical analysis of texting for classroom instruction. This is corroborated by Bernard (2008) whose study of texting in classrooms throws up widening opportunities for using this technology for comprehension purposes, to get a quick grasp of course material and a way to deconstruct and decipher sophisticated texts. Her example of using the entire Shakespearean play Richard III to be written in texting language may be viewed as creating a challenge which would test some of these skills.

Englishes in Literature While the long term sustainability of these strategies of combining literature and technology may be suspect, they do point to the very real possibilities of the way in which the borders of language studies have been pushed beyond the limitations of previously conceived margins. The theoretical overlapping of literature with technology is another manifestation of English being re-created and re-used by a widening group of writers for whom English is at once the language of choice but is also the only possible language to address a wider audience. Writing in English is very often a political statement, not bereft of associated problems such as loss of native languages. In a postcolonial, postmodern world, English has become central to the debate on identity and nation. To a large extent, however, the experiments of colonized nations with literary productions in

18

Chapter One

English tend to produce works which reflect on changing societies in a way that centres the experience of those margins. In the larger canvas, English literature has extended well beyond the limits of British and American writing to include, within its purview, a rapidly expanding range of writers from various corners of the world writing in varieties of English. This is largely because ideas of centres and margins have been gradually disappearing in the literary context owing to the difficulty of identifying specific communities in any particular geographical or literary tradition. This is nowhere more relevant than in present day diasporic communities where identities and languages coalesce into increasingly complex layers. Examples of multiple migrations include Indians who moved to Uganda in the early part of the twentieth century and then re-migrated to England following the political turmoil of the 1960s or of the migrations of Chinese who moved to neighbouring countries such as Hong Kong and subsequently migrated to America or Canada (Braziel and Mannur, 2003). The effect of these multiple migrations on the consciousness of the creative writer and the language in which they choose to write has been extensively explored (Rushdie, 1992; Loomba, 1998; Boehmer, 2005). The consequences of the linguistic choices made by such writers have been variously seen as being laudable (Chandra, 2000) or pandering to the international market (Mukherjee, 1996). In this sense, Arjun Appadurai’s notion that “the new global cultural economy has to be seen as a complex overlapping, disjunctive order, which can no longer be understood in terms of existing center-periphery models” (Appadurai, 1989, p. 324) is particularly apt in diasporic literatures. Which language to use in writing the postcolonial tale was an issue widely discussed and debated upon by a number of creative writers of the post-independence era of the 1950s. Their voices were famously represented by the Kenyan writer Nguigi who, in his Decolonising the Mind (1986), argued that English could no longer remain the primary medium of expression and that continuing to use English would be a continuation of colonialism. Nguigi’s stand, while being lauded by a number of regional writers, had to be compromised upon, as the realities of expressing and marketing within a local market could not achieve the kind of international recognition sought by a majority of writers. Salman Rushdie joined this debate and became one of the many ‘metropolitan’ writers to choose to write in English. As he explains: Those of us who do use English do so in spite of our ambiguity towards it, or perhaps because of that, perhaps because we can find in that linguistic struggle a reflection of other struggles taking place in the real world,

English Language Studies in the New Millennium

19

struggles between the cultures within ourselves and the influences at work upon our societies. (Rushdie, 1992, p.17)

With the language wars of newly independent countries of Africa and the Indian subcontinent having been largely settled, the focus in postcolonial literary studies has shifted to the way in which English continues to be changed in the writings which emerge from these parts of the world as well as from the immigrants who have taken their language to a predominantly English setting in the west. In significant ways, Chinua Achebe, Salman Rushdie, Wole Soyinka and Anita Desai have used language to effectively comment on, and transform the imperialist perception of English by accepting and using it in uniquely different ways, re-creating literary genres and transforming the language. It is not an easy struggle, as Rushdie presents through an internal dialogue in his novel Shame, in a conversation revolving around the idea of whose histories and whose languages must be privileged in a narrative: Outsider! Trespasser! You have no right to this subject!...I know; nobody ever arrested me. Nor are they ever likely to. Poacher! Pirate! We reject your authority. We know you, with your foreign language wrapped around you like a flag. Speaking about us in your forked tongue, what can you tell but lies? I reply with more questions: Is history to be considered the property of the participants only? In what courts are such claims staked, what boundary commissions map out the territories? Can only the dead speak? (Rushdie, 1983, p. 184)

The dilemma of the narrator above points to a major issue in postcolonial approaches revolving around the question of whose stories must be told and in whose language. Bill Ashcroft, Gareth Griffiths and Helen Tiffins in their iconic book on postcolonialism The Empire Writes Back (1989) suggest that these tensions regarding English as used by post colonial writers are solved by writers in two ways, either by ‘abrogation’ or ‘appropriation’ (p. 38) with the former being rejected as being an imposition of language over a former colony and the latter being a way in which a language (in this case, English) could be used by a society as a medium to express divergent interests and concerns. Increasingly, as Ismail S. Talib (2002) and other critics like Afzal Khan (2000) point out, literatures beyond English need to be brought to light and contextualized outside the paradigms as established by English literature but, as long as English remains a potent language, writers use it, transform it and in fact, re-form it to adapt to local contexts and audiences. The spin-off of these

20

Chapter One

varieties is being increasingly seen in humanities departments across the world shifting from using the label of ‘English Literature’ to ‘World Literature’ (Damrosch, 2006). It is thus inevitable that this shift in English usage would have wider implication for the way in which it is being taught in the classroom, along with associated problematics of accents, choice of vocabulary and assessment.

English in the Classroom The expanding sphere of English speaking communities and its various manifestations has had a tremendous impact on the methodology of English language teaching in the classroom. Teaching itself has become a very carefully used term, with many writers preferring more neutral terms such as ‘English in the classroom’. Following Foucault’s work on power and knowledge (Foucault, 1980), the pedagogical concerns of teaching English in classrooms have increasingly been seen as expressions of power and establishing personal agendas. Alaistair Pennycook’s notion that schools cannot be seen as neutral sites where a curricular body of information is passed on to students (Pennycook, 1994) has been taken up by a variety of researchers whose primary attempt is to dismantle power equations within the classroom and create possibilities for learners across the margins to participate in the learning process. Much of the research on ELT is presently guided by Braj Kachru’s classification of English speakers in expanding circles, as opposed to the linear division of centres and margins. Shifting from the late twentieth century focus on ‘communication competence’ (Hymes, 1972, p. 269) contemporary studies focus more firmly upon issues pertaining to varieties of English. Such studies include those of J.L. Dillard (1972), Robert Phillipson (1992), Michael Clyne (1992), Peter Strevens (1992) and Plat, Weber and Ho (1984) among many others. Many studies have used the theoretical implications of these works to relate them to ELT classrooms in terms of speaking, writing and assessment. Many of the areas of ELT have been considerably explored, including the central concern of using canonical English texts to teach English at the ‘peripheries’. Issues such as the place of world English in the English classroom (Peterson and Coltrane, 2003) and those of language competence among teachers, particularly the distinction between “genetic nativeness” and “functional nativeness” (Kachru, B. 2005, p. 12) have continually been addressed with the intention of arriving at common areas of agreement. These issues have even been institutionalized, particularly by agencies such as Cambridge ESOL which has gradually begun to recognize widening concerns

English Language Studies in the New Millennium

21

surrounding issues of accents, choice of vocabulary and sentence structure, choosing to focus instead, on organization, clarity of ideas and coherence as primary points of correction (Taylor, 2006).

Conclusion It can be safely assumed that the narrative of millennial language studies firmly focusses on the multiple identities of the various communities who use English in increasingly diverse ways. The repercussions of these varieties remain enormous and the possibilities endless. However, certain pointers may help to situate the discussion more firmly. From a methodological point of view, David Crystal’s comments in a conference of the IATEFL in 2001 are still relevant. Speaking of a 'nightmare' in which he would have to teach 15 year old students, he says, “I would want to be giving them a take on the language which was more compatible with their own age...and that would mean my taking into account the usage trends around them”. This approach, coupled with subsequent research on the power of the young to affect changes to language through changing identities and cultural perceptions could indicate the way in which language studies could set the agenda for future research by focusing on the vibrant community of English learners and users. Talbot et al (2003) dwell extensively on language changes affected by the youth, the young being defined as being a fluid range of individuals who move in and out of multiple contexts and identities: Today, research on ‘youth’–and accompanying definitions of what it means to be ‘young’–has moved on. Earlier, inflexible definitions have been replaced largely by approaches that take into account the new formations of youth culture as they are actively constituted in late modern society. There is an appreciation that young people draw on a range of heterogeneous and fluid cultural practices to articulate their multiple identities. (Talbot et al, 2003, p. 202)

English studies in the twenty-first century will have to continue to address the issues of how language is owned, by whom and who is transforming it. While it may continue to move across borders in the real and virtual world, English itself remains a contentious site for explorations and research. In a very real sense, “English as a field of knowledge now belongs to those who know it and use it” (Kachru, and Smith, 2008, p.180).

22

Chapter One

Bibliography Afzal-Khan, F. and K. Seshadri-Crooks. 2000. The Pre-Occupation of Postcolonial Studies. Durham: Duke University Press. Appadurai, Arjun. 1989. Disjuncture and Difference in the Global Cultural Economy. In Lechner Frank J. and John Boli, (eds). The Globalization Reader. London: Blackwell Publishers, 2000. pp. 322-330. Aschroft, Bill, Gareth Griffiths and Helen Tiffins. 1989. The Empire Strikes Back: Theory and Practice in Post-Colonial Literatures. New York: Routledge. Baron, Naomi. 2008. Always on: Language in an Online and Mobile World. London: Oxford University Press. Bernard, Sara. 2008. Zero Thumb Game: How to Tame Texting. Accessed 21 Sept. 2010 from http://www. edutopia.org/text-messaging-teachingtool. Boehmer, Elleke. 2005. Colonial and Postcolonial Literature: Migrant Metaphors. Oxford: Oxford University Press. Braziel, Jana Evans and Anita Mannur, (eds). 2003. Theorizing Diaspora: A Reader. MA: Wiley-Blackwell. Brown-Owens, A, M. Eason and A. Lader. 2003. What Effect does Computer-Mediated Communication, Specifically Instant Messaging have on 8th Grade Writing? Accessed Sept. 15, 2011 from http://webarchive.org/web/20030821214021/http://www.usca.edu/med tech/courses/et780/may03/groupsproject/cmc/-im.html. Carroll, Lewis. 1865. Alice in Wonderland. Edmund R. Brown (ed), 1932. Boston: International Pocket Library. Chandra, Vikram. 2000. The Cult of Authenticity. The Boston Review. Accessed 17 Sept.2010 from http://bostonreview.net/BR25.1/ chandra.html Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge, MA: The MIT Press. Clyne, Michael (ed). 1992. Pluricentric Languages: Differing Norms in Different Nations. Berlin: Walter de Gruyter Crystal, David. 1997. English as a Global Language. Cambridge: Cambridge University Press. —. 1980. Introduction to Language Pathology. London: Edward Arnold Ltd. —. 2001. Twenty-first Century English. Keynote speech to IATEFL annual conference. Accessed Sept 20, 2010 from http://www.davidcrystal.com/DC_articles/Internet10.pdf. —. 2008. Txting: the gr8 bd8. Oxford: Oxford University Press.

English Language Studies in the New Millennium

23

—. 2010. Interview with the BBC. Accessed Sept. 12, 2010 from http://www.bbc.co.uk/news/technology-10971949. Damrosch, David. 2006. What is World Literature? New Jersey: Princeton University Press. De Castell, S. 1996. On Finding One’s Place in the Text: Literacy as a Technology of Self formation. In W. F. Pinar (ed). Contemporary Curriculum Discourses: Twenty Years of JCT. New York: Peter Lang, pp. 398-411. Dillard, J.L. 1972. Black English: Its History and Usage in the United States. New York: Random House. Durant, Alan and Marina Lambrou. 2009. Language and Media. London: Routledge. Fairclough, N. 1989. Language and Power. London: Longman. Flescher, J. 1969. The Language of Nonsense in Alice in Wonderland. Yale French Studies, No. 43. Yale: Yale University Press. Accessed Sept. 17 2010 from http://www.jstor.org/stable/292964. Foucault, Michel. 1980. Power/knowledge: Selected Interviews and Other Writings 1972-1977. New York: Pantheon. Friedman, Thomas. 2005. The World is Flat: A Brief History of the Twenty-first Century. New York: Farrar, Straus and Giroux. Graddol, David. 1997. The Future of English–A Guide to Forecasting the Popularity of the English Language in the 21st Century. United Kingdom: The British Council. Graddol, D., J. Cheshire and J. Swan. 1994. Describing Language. Buckingham: Open University. Herring, S. C. 2000. Gender Differences in CMC: Findings and Implications. Computer Professionals for Social Responsibility Journal. 18 (1). Accessed Sept. 12, 2011 from http://archive.cpsr.net/ publications/newsletters/issues/2000/winter2000/herring.html. Huffaker, David A. and Sandra L. Calvert. 2005. Gender, Identity and Language use in Teenage Blogs. Journal of Computer-Mediated Communication, Vol. 10, issue 2. DOI: 10.11111j.1083-61012005.tb.00238.x. Hymes, D. 1972. On Communication Competence. In Pride, J. B. and J. Homes (eds). Sociolinguistics. Haemondsworth: Penguin Books. pp. 269-293. Jenkins, Jennifer. 2003. World Englishes: A Resource Book for Students. New York: Routledge. Kachru, Braj (ed). 1982. The Other Tongue: English across Cultures, rpt. 1992. Chicago: University of Illionis Press.

24

Chapter One

Kachru, Braj. 2005. Asian Englishes: Beyond the Canon. Hong Kong: Hong Kong University Press. Kachru, Yamuna and Larry E. Smith (eds). 2008. Cultures, Contexts and World Englishes. New York: Routledge. Lee, J. 2002, Sept 19. I think, therefore IM. New York Times, P.G.1. Lenhart, A., S. Arafeh, A. Smith and A. MacGill. 2008. Writing, Technology and Teens. Accessed Sept 25, 2010 from http://www.pewinternet.org/Pdfs/PIPWritingReport FINAL3:pdf. Loomba, Ania. 1998. Colonialism-Postcolonialism. New York: Routledge. Merchant, Guy. 2001. Teenage girls and Internet chatrooms. Accessed Oct. 20 2010 from http://www.shef.ac.uk/content/1/c6/05/05/23/merchant/pdf. Mukherjee, Meenakshi and Harish Trivedi. 1996. Interrogating Postcolonialism. Shimla: Institute of Advanced Studies. Nelson, Cecil L. And Braj B. Kachru. 1996. World Englishes. Sociolinguistics and Language Teaching. S. McKay and N. Hornberger (eds). Cambridge: Cambridge University Press. Nguigi, waThiong'o. 1986. Decolonising the Mind: The Politics of Language in African Literature. London: James Currey. Palmer, Harold E. 1924. Selected Writings. Richard C. Smith (ed). 1999. Tokyo: IRLT. Pennycook, Alaistar. 1994. The Cultural Politics of English as an International Language. Essex: Longman Group Ltd. Peterson, Elizabeth and Bronwyn Coltrane. 2003. Culture in Second Language Teaching. CAL Center for Applied Linguistics. Accessed Sept 17, 2010 from http://www/cal.org/resources/Digest/0309peterson.html. Piaget, J. 1971. Biology and Knowledge. Chicago: University of Chicago Press. Pinker, Steven. 1984. Language Learnability and Language Development. Cambridge MA: Harvard University Press. Phillipson, Robert. 1992. Linguistic Imperialism. Oxford: Oxford University Press. Platt, J., H. Weber and L. M. Ho. 1984. The New Englishes. London: Routledge. Rettberg, Jill Walker. 2008. Blogging. Digital Media and Society Series. Cambridge: Polity Press. Rushdie, Salman. 1983. Shame. London: Vintage. —. 1992. Imaginary Homelands: Essays in Criticism 1981-1991. London: Penguin Classics.

English Language Studies in the New Millennium

25

Savicki, V. 1996. Gender, Language Style and Group Composition in Internet Discussion Groups. Journal of Computer-Mediated Communication. 2 (3). Accessed Sept. 12, 2011 from http://ijcmc. indiana.edu/vol2/issue3/savicki.html Singhal, Meena. 1997. The Internet and Foreign Language Education: Benefits and Challenges. The Internet TEFL Journal. Accessed Sept. 17, 2010 from http://iteslj.org/Articles/Singhal-Internet.html. Skinner, B. F. 1957. Verbal Behavior. New York: Appleton. Strevens, P. 1992. English as an International Language. In Braj Kachru (ed). The Other Tongue: English Across Cultures. 2nd edition. Urbana: University of Illionis Press, pp. 27-47. Swales, J. 1990. Genre Analysis: English in Academic and Research Settings. Cambridge: Cambridge University Press. Talbot, M., K. Atkinson and D. Atkinson. 2003. Language and Power in the Modern World. AL: University of Albany Press. Talib, Ismail S. 2002. The Language of Post-Colonial Literatures. New York: Routledge. Taylor, Lynda. 2006. The changing landscape of English: Implications for Language Assessment. ELTJ. Accessed Sept. 12, 2011 from oxfordjournals.org/content/ 60/1/51.short. Tripathi, P.D. 1998. Redefining Kachru’s ‘Outer Circle’ of English. English Today. Cambridge: Cambridge University Press. Doi: 10.017/S0266078400010555 Van Dijk, T.A. (ed). 1985. Discourse and Communication: New Approaches to the Analysis of Mass Media Discourse and Communication. Berlin: de Gruyter. Vosloo, Steve. 2009. The effects of texting on Literacy: Modern Scourge or Opportunity? An issue paper from the Shuttleworth Foundation. Accessed Sept. 17, 2010 from citeseerx.ist.psu.edu/iewdoc/download. Vygotsky, L.S. 1962. Thought and Language. Cambridge, MA: MIT Press. Warschauer, M. 2001. Online Communication in Carter, R. and D. Nunan, (eds). The Cambridge Guide to Teaching English to Speakers of Other Languages. Cambridge: Cambridge University Press, pp. 207-212. Wright, Scott and John Street. 2005. Democracy, Deliberation and Design: The Case of Online Discussion Forums. New Media Society. Vol. 9 (5). pp. 849-869. Accessed Sept. 14, 2011 from http://nms.sagepub. com/content/9/5/849.full.pdf. Yano, Y. 2001. World Englishes in 2000 and Beyond. World Englishes. Vol. 20, Issue 2, 119-132. Doi: 10.1111/1467-971X.00.204.

CHAPTER TWO LANGUAGE AND GROUP IDENTITY: SOME SOCIAL PSYCHOLOGICAL CONSIDERATIONS ITESH SACHDEV

Abstract The last few decades have seen an exponential increase in worldwide mobility and exchange of people, ideas, money, food, etc., that has radically stretched the boundaries of societies by making them more diverse than ever in terms of ethnicities, cultures, languages, and religions. The main aim of this paper is to use a variety of empirical illustrations from social psychological research to explore how language reflects and creates group identity.

Keywords: identity, language, intergroup, minority, attitudes, indigenous peoples

Introduction Persons belonging to national or ethnic, religious and linguistic minorities…have the right to enjoy their own culture, to profess and practice their own religion, and to use their own language in private and in public, freely and without interference or any form of discrimination. (Article 2:1, United Nations Declaration on the Rights of Persons Belonging to National or Ethnic, Religious and Linguistic Minorities, 1992)

The last few decades have seen a stretching of local, regional and national identities, cultures, and economies worldwide as a consequence of

Language and Group Identity: Some Social Psychological Considerations 27

globalisation and technology development. There appears to have been an exponential increase in worldwide movement and exchange of people, ideas, materials, money, food and so on (but see Ghemawat, 2011). This has quite literally ‘stretched the boundaries’ and radically increased the complexity and make-up of modern societies such that there is greater than ever diversity in terms of ethnicities, cultures, and religions. An important aspect of this diversity concerns linguistic variation. Languages, accents, dialects and other linguistic variations not only provide important cues for the categorization of speakers on the socio-cultural map but can also emerge as the most salient dimensions of group identity (Fishman, 1977; Giles, Bourhis and Taylor, 1977; Sachdev and Bourhis, 2005). The main aim of this chapter is to explore the complexity of the multi-factorial relationship between language and group identity, given the importance of language as a dependent and an independent variable in analyses of relations between and within groups. The first section provides an illustration of how actual linguistic exchanges in the briefest of interactions may reflect and create significant intergroup convergence and differentiation. The second section illustrates how self-reported group identifications influence patterns of language use and attitudes. In the third section, the relative influences of cultural, religious and national identifications in the maintenance of minority linguistic identity in one modern diaspora are considered. All three sections report specific empirical illustrations employing social psychological methods and concepts. They are by no means exhaustive or comprehensive, but serve to illustrate the key point that the relationship between language and identity is a dynamic one, one which varies primarily as a function of the power relationship between groups (Sachdev and Bourhis, 1990).

Intergroup Convergence and Differentiation in Actual Language Use Bilinguals and multilinguals constitute a majority of people in the world (Baker and Jones, 1998; Edwards, 1994; Hamers and Blanc, 2000; Sachdev and Bourhis, 2001). Communication where two or more languages (and dialects) are used, generally involves speakers of different groups (Johnson, 2000; Sachdev and Giles, 2004). Which languages are used, when, why, and by whom are important questions given the crucial role that languages play not only in fostering effective intergroup communication, but also in terms of identification and solidarity (Fishman, 1977; Sachdev and Giles, 2004). The variety of interethnic contexts, coupled with the equally diverse array of ethnolinguistic behaviours and

28

Chapter Two

attitudes, represent a considerable challenge to modelling the relationship between bilingual/multilingual communication and social identifications. Recent models, described in detail elsewhere (e.g., Sachdev and Bourhis, 2001; Sachdev and Giles, 2004), specify outcomes in terms of behaviour and attitudes (including discourse), both in the immediate interpersonal interactional contexts, and also across individuals in terms of their acculturation (e.g., as a consequence of mobility), across societies in terms of multiculturalism and multilingualism, and across time in terms of intergenerational shift and maintenance. Predictors of these outcomes are generally classified at macro (e.g., dominant ideologies, policies and laws; group numbers, power, status) and micro (e.g., immediate sociolinguistic context norms and rules, networks of linguistic contact) levels which are thought to be mediated by social psychological factors (e.g., identities, attitudes, beliefs, perceptions). Since these models are detailed elsewhere the following is a brief empirical illustration of how language can be both a dependent and an independent variable in analyses of actual communication between and within groups. Studies conducted by Lawson-Sako and Sachdev (1996) and Lawson and Sachdev (2000) in Tunisia provide interesting material for this illustration, and a small set of their findings is referred to here. Tunisia’s position in the southern Mediterranean, in northern Africa, at the intersection between Europe, Africa and the Middle East, has made it very important strategically. Consequently, over the ages, originating with the Amazigh (‘Berber’) many different groups including the Phoenicians, Romans, Vandals, Arabs, Turks and French have contributed to its rich ethnolinguistic heritage. Perhaps the most enduring influence has stemmed from the Arabic invasion from the east in the seventh century (Hourani, 1991). Today a majority of the population has some Arab ancestry, the official religion of modern Tunisia is Islam, and Article 1 of the Tunisian Constitution states that the official language of Tunisia is Arabic. Arabic served as a powerful symbol of Tunisian identity in the independence struggle against the French colonisers (1881-1956), who had followed more than three hundred years of Turkish rule (Maamouri, 1983). The Arabic spoken in non-formal contexts is Tunisian Arabic (TA), which is in a diglossic (Ferguson, 1959) relationship with Modern Standard Arabic (MSA). Alongside Arabic, the ex-colonial language, French, has continued to maintain a prominent place in the Tunisian linguistic landscape, and French is used extensively enough for LawsonSako and Sachdev (1996) to refer to it as having high linguistic 'vitality' in Tunisia (Sachdev and Bourhis, 1993). As a result of the high vitality (and

Language and Group Identity: Some Social Psychological Considerations 29

actual frequency of use) of both Arabic and French, Tunisia is frequently characterised as a bilingual country. Lawson-Sako and Sachdev (1996) and Lawson and Sachdev (2000) conducted a series of studies in the Tunisian context examining language behaviour and attitudes. They focused one of their studies on actual language behaviour in the street to maximise the external validity of their findings as the street constitutes a relatively anonymous environment where people may be inclined to act ‘naturally’. They employed an experimental methodology to see how the ethnic backgrounds of Tunisian researchers (‘brown’ Arab, ‘white’ European and ‘black’ African) and the language (Tunisian Arabic/French) they used to make a request for directions, was related to the language of responses by a large sample (1000+) of randomly selected Tunisian pedestrians of Arab origin. In this context, assuming that bilingual competence is not an issue, one may expect a variety of different outcomes depending on the effects of macro and micro variables as mediated by social psychological factors. For instance, some may suggest that people are more likely to respond in Arabic as it is the official language (macro level effect); others may suggest that pedestrians are more likely to respond in the language in which the request was posed i.e., ‘converge’, as this would be an appropriate micro-sociolinguistic norm and/or because it is more helpful and accommodating (see Sachdev and Giles, 2004); yet others may come up with other complex hypotheses that suggest an interaction of various different factors. Lawson-Sako and Sachdev’s (1996) findings showed that Tunisian pedestrians were extremely accommodating and cooperative in fulfilling the relatively innocuous and short-lived requests for assistance. Overall, a majority of pedestrians converged by responding in the language of the question providing support for micro-normative factors or social psychological factors concerned with communication accommodation (see Sachdev and Giles, 2004). However, the overall pattern masked systematic variation. For instance, they found that whereas significant amounts of responses to ‘brown’ Tunisians were in a mixture of Arabic and French (they refer to it as “code-switching”), little or no ‘code-switching’ took place in response to questions from their ‘white’ or ‘black’ compatriots, regardless of the language they used to ask the question. Lawson and Sachdev (2000) suggest that ‘code-switching’ is an important aspect of Arab Tunisian ingroup identity, and has the “ability simultaneously to connote both status (via French) and solidarity (via Tunisian Arabic)”. Indeed, there is an increasing body of evidence that suggests that language mixing is a key characteristic of communication in bilingual and

30

Chapter Two

multilingual contexts (Cheshire and Gardner-Chloros, 1996; Sachdev and Bourhis, 2001, also see Garcia’s, 2007, ‘translanguaging’). Clearly, given that multilingualism is normative worldwide, the manner, role, motivations, attitudes and identifications associated with using/mixing several languages in conversation merits further study. Questions posed by male ‘white’ researchers in Lawson-Sako and Sachdev (1996) received vast amounts of convergence, regardless of the language of the question, but with ‘black’ male questioners, almost half of the respondents diverged by using a language other than the one the question was posed in. According to the Bilingual Accommodation Model (Sachdev and Giles, 2004), divergence is a function of a speaker's desire to dissociate from the interlocutor, while convergence suggests pro-social orientations. As Lawson-Sako and Sachdev (1996, pp.75-76) wrote: Evidently, Tunisian subjects in this study exercised a real choice about which language to use and were not constrained by questions of linguistic competence. Subjects' linguistic choices were identity choices, varying systematically as a function of the language, ethnicity… of the researcher…. permeating even the briefest interactions in multilingual settings such as the streets of Tunisia.

Group Identification, Language Use and Attitudes In discussing the findings of the field experiment above, inferences were made about social categorisation and group identification through the observed patterns of systematic language variation, though no measures of social identity were actually obtained. Group identity is a complex matter and how group members label themselves and are labelled by others varies considerably. Labels chosen by group members to describe themselves facilitate identity formation, as well as providing the vehicle for experiencing and communicating their meaningful identities (Sachdev, Arnold and Yapita, 2006; Larkey, Hecht and Martin, 1993). The focus of this section is on an empirical illustration of the association of identity labels with patterns of the use and attitudes amongst some indigenous languages in the Americas (see detailed report by Sachdev et al, 2006). Indigenous languages and cultures flourished in the lands now known as the Americas before the arrival of Cristoforo Colombo in 1492. Indigenous peoples were the majority of the population in the Americas probably until the middle of the 19th century, and their languages were the main vehicle for articulating and transmitting their cultures, identities and spiritualities. European expansion involving warfare, the spread of disease, land alienation, displacement, isolation in reservation systems, enforced

Language and Group Identity: Some Social Psychological Considerations 31

assimilation and various other governmental measures including largescale non-indigenous immigration, led to a situation where indigenous peoples comprise small minorities in most of the Americas today, except in the Andean region (e.g., Bolivia). Throughout the Americas, European colonisation aimed to totally assimilate indigenous peoples, eradicate indigenous languages and cultures, and deny them any vitality as distinctive groups or nations. It was accompanied by a powerful racist ideology which characterised such peoples as ‘uncivilised’ and ‘barbarian’. Colombo had sailed west in the 15th century from Europe looking for Asia, and landed in the Antilles (in the Caribbean) thinking he had arrived in the islands known to the Europeans as the ‘Indies’ in the Indian Ocean. He mistakenly labelled the inhabitants ‘Indios’ (‘Indians’), and even though Colombo's mistake was soon recognized, this colonial label (and variants of it such as ‘American Indian’, ‘Native Indian’) has proved quite ‘resilient’ over centuries in spite of efforts by some groups to rename themselves in order to reclaim their distinctive identities. For instance, Sachdev et al (2006) discuss how the label ‘Indian’, to refer to those of indigenous origin, has been a mainstay of official government discourse in Canada, is internalised as a preferred label by some indigenous communities in the USA, yet discarded in favour of the label ‘First Nations’ by a large number of indigenous communities in Canada. The specific labels compared by Sachdev et al (2006) in a quantitative correlational study were those associated with European colonisation (“Indian” in Canada, and “Indio” in Bolivia), and those associated with postcolonial indigenous identities of the specific groups in the research (“Cree”/“Haida” in Canada, and “Aymara” in Bolivia). Employing a questionnaire methodology with a moderate-sized sample (under 200) of indigenous adults and adolescents, self-reports of their identifications with colonial and postcolonial group labels as well as their proficiencies, use and attitudes concerning own-group (“Cree”/“Haida”/“Aymara”) and colonial languages (English in Canada/Spanish in Bolivia) were obtained. Overall, Sachdev et al’s (2006) findings showed that identification with European colonial group labels (“Indian”/“Indio”) correlated positively with the use of (and attitudes towards) European languages (English/Spanish), and negatively with the use of (and attitudes towards) indigenous languages (Cree/Haida/Aymara). Conversely, identification with postcolonial labels (“Cree”/“Haida”/“Aymara”) was related positively with the use of (and attitudes towards) indigenous languages (Cree/Haida/Aymara) and negatively with the use of (and attitudes towards) European languages (English/Spanish). Although this research

Chapter Two

32

was relatively small-scale and employed a somewhat limited quantitative methodology, the findings of this research demonstrate cogently how selfreported group identification–operationalised by a fairly straightforward and somewhat simplistic ingroup labelling process–relates systematically to language use and attitudes. Overall, the findings of this small-scale study reinforced the important notion of the link between group identity and language. From the point of view of the ethnolinguistic vitalities of indigenous groups in Bolivia and Canada, Sachdev et al (2006) argued that the state apparatus that emerged from colonial times had largely failed to address the legal status of indigenous languages and cultures, and the rights of indigenous peoples to define their own educational policy, materials and curricular development according to their own criteria. Today there are optimistic signs for indigenous languages in Bolivia, due to initiatives expected of their first-ever indigenous President—Evo Morales—and in Canada, due to some recent initiatives such as the Canadian Government’s Aboriginal Languages Initiative that began in 1998 and has renewed funding until 2014. However, it is important to note that the struggles for increased ethnolinguistic vitality of indigenous (and other minority) groups need to be contextualised within broader political, economic, environmental and socio-cultural relations of power between groups in societies. Let us now turn our attention from indigenous minority groups to modern immigrant minorities that are stretching linguistic and other boundaries in diaspora due to increased mobility and advances in information technology (Fenoulhet and Ros I Sole, 2011).

Minority Linguistic Identities in Diaspora In the summer of 1994, when I was visiting North America I came across a small advertisement in the classified section of India Abroad, a weekly newspaper distributed across the United States and Canada for Indians like me who live in diaspora (printed edition 10.6.94, it now also has an online edition at www.indiabroad.com). The advertisement began with a bold claim: Indian Accents Erased with this Powerful Audio-Visual tool (audio cassettes and book)

with the promise to ...erase the foreign accents of Indians who already know how to speak English.. and who desire to polish their communication skills by mastering

Language and Group Identity: Some Social Psychological Considerations 33 the color [my emphasis], sounds and inflection of standard American English!

I have presented this advertisement to a variety of conference audiences (national and international) while reading aloud the words in a stereotypical Indian accent to great amusement. Two weeks after the appearance of this advertisement, I visited Australia and came across another diaspora publication–the ‘Indian Down Under’ (today at www.indiandownunder.com), with an advertisement for “Formula 10 whitening cream” that promised to give the skins of darker hued Indian readers a “fairer complexion”. In this advertisement, although the focus was not linguistic (and perhaps less amusing), it would be fair to argue that the underlying implied social psychological process is not too different from that underlying the accent advertisement mentioned above. The messages of assimilation for minorities implied by these advertisements are not restricted to diaspora newspapers, and at the current juncture, at the beginning of the second decade of the 21st century, several western governments (Germany, UK, Italy, France, Denmark, The Netherlands) have public policies that aim to ‘integrate’ ethnolinguistic minorities by introducing measures that are actually designed to assimilate and devalue their heritage languages and cultures. Given that these official policies aim to change group identities, where does (do) language(s) fit in? Much social psychological research has suggested that language and identity appear to be related reciprocally: language use influences the formation of group/cultural identity and group/cultural identity influences patterns of language attitudes and usage (see above; Giles et al, 1977; Sachdev and Bourhis, 2005). A good qualitative example was provided by Paolillo (1996) who reported the following exhortation for the primacy of language over religion, while stressing the overarching importance of culture to language: To any Punjabi out there, whether you are a Hindu, Muslim or Sikh: If you speak Hindi(Urdu) instead of Punjabi, you are a serious disgrace to your culture and you shouldn't call yourself a Punjabi!

There is also systematic quantitative evidence from studies using multidimensional scaling that have shown ethnolinguistic group members identifying more closely with those who spoke their ‘native’ tongue than with those who shared their cultural background or geographic origin (Giles et al, 1977). Another example can be found in the revival of ancestral languages which, despite not being spoken by the majority of members of a group, often become a central issue around which group

34

Chapter Two

members mobilise to affirm or redefine their group identities (e.g. Fishman, 1999; Giles and Johnson, 1981). At a macro-societal level, language planning has long been used as a tool for nation building across the world (Bourhis, 1984; Fishman, 2001). For example, the spectacularly successful revival of Hebrew from being a ‘dead’ language just over half a century ago to the majority language in Israel today, has been directly attributed to the language planning conducted by the State of Israel in its formative stages (Spolsky and Shohamy, 1999). Evidence for the centrality of language to group identity may also be found in studies that show linguistic and cultural assimilation by members of minority groups adopting majority group languages and identities (Bourhis and Sachdev, 1984; Sachdev and Bourhis, 1990; see Blackledge and Creese, 2008 for an interesting sociolinguistic critique of essentialist approaches). For instance, Kanazawa and Loveday (1988), found that third generation immigrants of Japanese ancestry in Brazil, who had linguistically assimilated to the degree that they were monolingual in Portuguese, identified as Brazilian and not Japanese. The current proclamations by many European governments (e.g., by German and UK prime ministers in 2010 and 2011) about the ‘failure of multiculturalism’ and the necessity of learning national languages for citizenship provide further fuel for the identity-language link, especially as the public discourse of ‘integration’ appears to actually suggest assimilation since minority cultures and languages are not valued. In the USA, anglophone right-wingers have frequently opposed education in languages other than English with some even echoing the phrase often attributed to the early 20th century Texas governor, Miriam ‘Ma’ Ferguson: “If English was good enough for Jesus Christ, it's good enough for us"! The use by right wing lobbies of linguistically assimilated high status members of immigrant groups (such as Hayakawa) in the promotion of English as the only official language of the USA provides further interesting examples of the appeal to the intertwined relationship between identity and language (but see Barker and Giles, 2002). Relatedly, Garcia and Bartlett (2007) discuss how English (only) has recently replaced bilingual (Spanish-English) educational programmes and policies in the USA as the official dominant ideology and discourse. Dominant, official government policies and ideologies for minorities vary considerably across countries from, for instance, assimilation in France, through the civic ideologies such as those of the UK and the ‘melting pot’ of the USA, to the more pluralistic official multiculturalism (‘mixed salad’) of Canada and Australia (see Berry, 1974; Sachdev and

Language and Group Identity: Some Social Psychological Considerations 35

Bourhis, 2001, and Bourhis et al, 1997 for more elaborate discussion of integration, assimilation and other acculturation orientations). The role of minority language maintenance needs to be contextualised in these macrolevel policies. Unfortunately, cross-national studies that allow us to compare at the macro-level in this field are rare due to a host of complex logistical, pragmatic, conceptual and methodological difficulties associated with obtaining comparable samples from contexts that are economically, socially and culturally diverse. However, a useful illustration of minority group linguistic identity in diaspora across different nation states may be obtained from research by Sachdev (1995) that focused on Punjabi speaking Sikhs in India, UK and Canada. According to various official data and documents (including censuses where there are questions on language, e.g., India Census, 2001; Pakistan Census, 1998; Canada Census, 2006; but there has been no language question in UK until 2011 Census which is being conducted now) the number of speakers of Punjabi worldwide may be estimated to amount to more than 100 million speakers. The majority are to be found in Pakistan (over 70%), where Urdu is the official language. In India, Punjabi speakers, although a very small community forming less than four per cent of the current population of India, are in the majority (over 70 per cent) in the state of Punjab. The state of Punjab has a rich multilingual heritage and Punjabi, as the official state language, enjoys higher vitality than Hindi and English (which also have considerable prestige and official support in Punjab). Significant numbers of Punjabi speakers are also to be found in neighbouring states, including Delhi (the national capital), where Hindi speakers are in a majority and where Hindi has official and dominant language status. Hindi speakers dominate in Delhi demographically while their political power is shared with an elite English speaking minority. Punjabi, which has a relatively high demographic base in Delhi (estimates vary from 13 per cent to 25 per cent of population), is mainly restricted to informal non-official domains (Sachdev, 1995), although since the mid-1990s street names and some other official facilities are available in Punjabi, Hindi and English in Delhi. In the second half of the 20th century, Punjabi has also increasingly begun to occupy a visibly significant space on the linguistic landscapes of several countries outside of the Indian sub-continent including United Kingdom, Canada, USA, Singapore, Malaysia, New Zealand as well as countries in the Gulf and in East Africa (e.g., Agnihotri. 1987, Bhachu, 1986; Khuswant Singh, 1991).

36

Chapter Two

Brass (1994) discusses how linguistic identities in north-western India were shaped by conflict along political and religious dimensions in the 19th and 20th centuries. Vis-à-vis Punjabi, Sachdev (1995) summarises thus: ...the polarisation of religious differences in the Punjab, the bi-national 1947 partition of the region of Punjab and the further 1966 partition of the Indian state of Punjab into three separate entities divided along linguistic lines, left Sikhs as the prime 'owners' and promoters of the Punjabi language.

The Sikhs, a majority in Punjab (59% according to India Census 2001), are highly visible (though in a minority) in the larger urban centres of North India such as Delhi outside of the Punjab (the current Prime Minister of India, Dr Manmohan Singh, is a Sikh). For Sikh communities outside of the Punjab, in India and elsewhere (e.g., UK, USA and Canada), there is some evidence from studies conducted by different researchers in their own national contexts (e.g., UK, USA, India, etc, see Sachdev, 1995) that suggested that Sikhs display high levels of linguistic maintenance in spite of inhabiting national contexts where government policies encouraging linguistic and cultural diversity are few and far between (e.g., USA, UK), and where dominant languages like English are a powerful draw for immigrant communities (Agnihotri, 1987). Generally, Punjabi Sikh settlements are accompanied by the establishment of gurdwaras (places of worship) which serve as important foci for supporting the cultural, political, social (including welfare support) and religious lives of local Sikh communities. Moreover, almost all gurdwaras have some form of arrangements for the teaching of Punjabi (using the Gurmukhi script) to Sikh children and adults. Given that previous literature had suggested that there were a variety of factors involved in the maintenance of Punjabi linguistic identity including religious, cultural, national and the power of other dominant languages, and in the absence of appropriate comparative data, Sachdev (1995) designed a quantitative cross-national study to understand the resilience of Punjabi linguistic identity. The study was conducted amongst Sikh teenagers in 3 national contexts, India, Canada and UK and four urban centres: in Amritsar (in the state of Punjab), India (Punjabi dominant majority setting); in Delhi (the national capital), India (outside the Punjab, Hindi dominant, Punjabi minority setting in India); and also in the two of largest populations of Sikhs outside of India—in London, UK (English dominant, Punjabi minority outside of India) and in Vancouver, Canada (English dominant, Punjabi minority setting outside of India).

Language and Group Identity: Some Social Psychological Considerations 37

India may be considered one of the most favourable contexts for multilingualism with over 20 national languages which include Punjabi, and an official version of geo-linguistic language planning where most of the national languages are generally majorities in their home state (in this case, the state of Punjab), and where the school education system aims to promote the teaching and learning of three languages (but see critique by Mohanty, 2010). Canada also has a pluralistic approach encapsulated by an official policy of federal multiculturalism that is designed to operate through English and/or French as official languages, and as such does not provide much support for 'non-official' language maintenance. Education in Canada is under provincial jurisdiction, conducted through the medium of one (or more infrequently both) of the official languages. Recently (last decade or so), with the help of the BC provincial government, schools in places like Vancouver have begun to provide support for 'non-official' languages like Punjabi, though not without opposition from significant sections of the dominant group. In the UK, Punjabi language teaching has largely been in the hands of the family and the community. Government support for community languages, in a country where English is the official language (also Welsh in Wales), has generally been sparse and unsystematic in spite of repeated community demands (e.g., McPake and Sachdev, 2008). There has been some success in that a handful of schools in London and other parts of the UK now offer Punjabi as a subject on the curriculum. However, given the low status ascribed to community languages (McPake and Sachdev, 2008), Punjabi language teaching is set for an uncertain future in the UK. Unlike India and Canada, UK has no national policy supporting cultural pluralism in spite of the fact that all major urban centres and many UK citizens are multilingual and multicultural. Sachdev’s (1995) main findings were unambiguous in that identification as Punjabi Sikh was high and resilient, able, in large part to resist the stretching of regional and national boundaries due to migration and mobility. Importantly in all national contexts, Punjabi cultural identities were consistently and positively related to Punjabi linguistic identities, reinforcing the widely recognised view that language and culture are intricately intertwined. There was also good evidence that identification with Sikhism positively predicted Punjabi linguistic identity in northern India but not outside of India. Interestingly, and as might be expected from a macro-level analyses, national identities subtracted from linguistic identity outside of India but not in India. Employing Turner et al’s (1987) Self Categorisation Theory, Sachdev (1995) argued that religious selfcategorisations were extended to linguistic differentiation in India

38

Chapter Two

probably due to the historicity of important sociopolitical and religious differentiation that had taken place in the Punjab. The extension of these differentiations to incorporate linguistic differentiation serves to maximise the meta-contrast between Sikhs and others in India. In contexts outside of India in this study, it is likely that ethnicity rather than religion served to maximise the meta-contrast between ingroups and outgroups (in Canada and UK). Overall, the findings of this study suggest that the association between language and identity is cross-nationally robust, and that supportive national and regional policies (and their implementation) are important for the maintenance of minority linguistic identities.

Summary and Concluding Notes This paper focussed on stretching our understanding of the complex relationship between language and group identification. It began with a discussion of how identity processes permeate even the briefest of actual anonymous linguistic interactions with languages being employed to increase or decrease levels of intergroup differentiation. This was followed by consideration of evidence that directly linked self-reported group identifications to language use and attitudes amongst indigenous peoples. The complexities of minority linguistic identity were further explored in diaspora in order to assess the relative influences of dominant language, cultural, religious and national identifications. Several conclusions may be reached from this paper. Firstly, bilingual behaviour is multi-factorial, and in all likelihood dynamically evolves throughout the course of communication. Secondly, the use of multimethodological approaches that focus not only on actual language use, but also on beliefs, perceptions and attitudes, is crucial in understanding the relationship between language and identity. Thirdly, identity appears to be key for language maintenance (and even revitalization) with the emphasis being on self-determination and rejection of imposed categorizations. It is important to reiterate that the relationship between language and group identity is not static and varies as a function of the power relations between groups. The struggles for linguistic justice and rights by minorities may stretch existing boundaries, but they need to be contextualised as part of a broader empowerment of minorities.

Acknowledgements I would like to thank Dr Balasubramanian, Dr Andrew Littlejohn and their colleagues for their support and patience in the writing of this paper.

Language and Group Identity: Some Social Psychological Considerations 39

Bibliography Agnihotri, R. K. 1987. Crisis of identity: The Sikhs in England. Delhi: Bahri Publications. Baker, C. and S. Prys Jones. 1998. Encyclopaedia of Bilingualism and Bilingual Education. Clevedon: Multilingual Matters Barker, V. and H. Giles. 2002. Who Supports the English-only movement?: Evidence for misconceptions about Latino group vitality. Journal of Multilingual and Multicultural Development, 23, pp. 353370. Berry, J. W. 1974. Psychological aspects of cultural pluralism: Unity and identity reconsidered. Topics in Cultural Learning, 2, 239-252 Blackledge, A. and A. Creese. 2008. Contesting ‘language’ as ‘heritage’: Negotiation of identities in late modernity. Applied Linguistics, 29, pp. 533-554 Bhachu, P. 1986. Twice immigrants—East African Sikh settlers in Britain. London: Tavistock Press. Bourhis, R.Y. (ed). 1984. Conflict and language planning in Quebec. Clevedon: Multilingual Matters. Bourhis, R.Y. L.C Moise, S. Perrault and S. Senecal. 1997. Towards an interactive acculturation model: A social psychological approach. International Journal of Psychology, 32, pp. 369-386. Bourhis, R. Y. and I. Sachdev. 1984. Vitality perceptions and language attitudes: some Canadian data. Journal of Language and Social Psychology, 3, pp. 97-126. Brass, P. R. 1994. Language, Religion and Politics in North India. Cambridge: Cambridge University Press. Canada Census. 2006. Statistics Canada–2006 Census Language, http://www12.statcan.ca/census-recensement/2006/rt-td/lng-eng.cfm (accessed 4/5/2011). Cheshire, J. and P. Gardner-Chloros. 1996. Code-switching and the sociolinguistic gender pattern. International Journal of the Sociology of Language, 129, pp. 5-34. Edwards, J. 1994. Multilingualism. London: Routledge. Fenoulhet, Jane and Cristina Ros i Solé (eds). 2011. Mobility and Localisation in Language Learning: A View from Languages of the Wider World. Oxford: Peter Lang. Ferguson, C. A. 1959. Diglossia. Word, 15, pp. 325-40. Fishman, J. A. 1977. Language and ethnicity. In H.Giles (ed). Language, Ethnicity and Intergroup relations, pp.15-57. London: Academic Press.

40

Chapter Two

—. 1999. Handbook of language and ethnicity (ed). New York: Oxford University Press. Fishman, J. A. (ed). 2001. Can Threatened Languages be Saved? Reversing Language Shift, Revisited: A 21st Century Perspective. Clevedon: Multilingual Matters. García, O. 2007. Foreword. In S. Makoni and A. Pennycook (Eds.). Disinventing and Reconstituting Languages, pp. xi–xv. Clevedon, UK: Multilingual Matters. Garcia, O. and L. Bartlett. 2007. Educating Speech Communities: An Unusual School Model for Latino Newcomers in an Era of Standards. International Journal of Bilingual Education and Bilingualism, 10, pp. 1-25. Giles, H. and P. Johnson. 1981. The role of language in ethnic group relations. In J.C. Turner and H. Giles (eds). Intergroup Behavior, pp. 199-243. Oxford: Blackwell. Giles, H., R. Y. Bourhis and D. Taylor. 1977. Towards a theory of language in ethnic group relations. In H. Giles (ed). Language, Ethnicity and Intergroup Relations, pp. 307-48. London: Academic Press. Ghemawat, P. 2011. World 3.0: Global Prosperity and How to Achieve It. Boston, USA: HBR press Hamers, J. F., and M. H. A. Blanc. 2000. Bilinguality and Bilingualism. Cambridge: Cambridge University Press. Hourani, A. 1991. A History of the Arab Peoples. London: Faber and Faber India Census. 2001. Census: data on language, http://www.censusindia.gov.in/Census_Data_2001/Census_Data_Onli ne/Language/data_on_language.html, (accessed 4/5/2011). Kanazawa, H. and L. Loveday. 1988. The Japanese immigrant community in Brazil: Language contact and shift, Journal of Multilingual and Multicultural Development, 9, pp. 423-435. Khushwant Singh. 1991. A History of the Sikhs, Volumes 1 and 2. Oxford: Oxford University Press. Johnson, F. (2000). Speaking Culturally: Language Diversity in the United States. Thousand Oaks: Sage. Larkey, L. K., M. Hecht and J. Martin. 1993. What’s in a name? African American ethnic identity terms and self-determination. Journal of Language and Social Psychology, 12, pp. 302-317. Lawson-Sako, S. and I. Sachdev. 1996. Ethnolinguistic Communication in Tunisian Streets, In Y. Suleiman (ed). Language and Ethnic Identity in the Middle East and North Africa, pp. 61-79. Richmond: Curzon Press.

Language and Group Identity: Some Social Psychological Considerations 41

Lawson, S., and I. Sachdev. 2000. Codeswitching in Tunisia: Attitudinal and behavioral dimensions. Journal of Pragmatics, 32, pp. 1343-1361 Maamouri, Mohamed . 1983. The Linguistic situation in independent Tunisia. In R.M. Payne (ed). Language in Tunisia, pp. 11-21. Tunis: The Bourguiba Institute of Modern Languages. McPake, J., and I. Sachdev. 2008. Community Languages in Higher Education: Towards realising the potential. LLAS ‘Routes into Languages’ research report at http://www.routesintolanguages.ac.uk /community (accessed 4/5/2011). Mohanty, A. K. 2010. Languages, inequality and marginalization: implications of the double divide in Indian multilingualism. International Journal of the Sociology of Language, 2010 (205), pp. 131-154. Paolillo, J. C. 1996. Language choice on SOC.CULTURE.PUNJAB, The Electronic Journal of communication, 6, http://www.cios.org/ EJCPUBLIC/006/3/006312.HTML (accessed 4/5/2011). Pakistan Census. 1998. Population Census. http://www.census.gov.pk/Statistics.htm (accessed 4/5/2011). Sachdev, I. 1995. Predicting Punjabi linguistic identity: From high to low in-group vitality contexts. International Journal of Punjab Studies, 2, pp. 175-194. Sachdev, I., D. Arnold and J. de Dios Yapita. 2006. Indigenous identity and language: some considerations from Bolivia and Canada, some Empirical Data. Birkbeck Studies in Applied Linguistics, 1, 81-102, http://bisal.bbk.ac.uk/ (accessed 4/5/2011). Sachdev, I., and R. Y. Bourhis. 1990. Language and social identification. In D. Abrams and M. Hogg (eds). Social Identity Theory: constructive and critical advances, pp. 101-124. Hemel Hempstead: HarvesterWheatsheaf. Sachdev, I. and R. Y. Bourhis. 1993. Ethnolinguistic vitality and social identity. In D. Abrams and M. Hogg (eds). Group motivation: social psychological perspectives, pp. 33-51. Hemel Hempstead: Harvester Wheatsheaf. Sachdev, I., and R. Y. Bourhis. 2001. Multilingual communication. In P. Robinson and H. Giles (eds). The New Handbook of Language and Social Psychology, pp. 407-428. Chichester and New York: Wiley and Sons. Sachdev, I. and R. Y. Bourhis. 2005. Multilingual communication and social identification. In J. Harwood and H. Giles (eds). Intergroup Communication: Multiple Perspectives, pp. 65-92. New York: Peter Lang.

42

Chapter Two

Sachdev, I., and H. Giles. 2004. Bilingual accommodation. In T. Bhatia and W. Ritchie (eds). The Handbook of Bilingualism, pp. 353-378. Oxford: Blackwell. Spolsky, B. and E. Shohamy. 1999. The Dynamics of Israeli Language Policy.Washington, DC: The National Foreign Language Center. John Hopkins University. Turner, J.C., M. Hogg, P. Oakes, S. Reicher, and M. Wetherell. 1987. Rediscovering the Social Group: A Self-Categorization Theory. Oxford: Blackwell.

CHAPTER THREE PROCEDURES FOR TRANSLATING CULTURALLY SPECIFIC ITEMS JAMES DICKINS

Abstract The translation of items (words and phrases) which are specific to one culture from a Source Language expressing that culture (the Source Culture) into a Target Language expressing another culture (the Target Culture) necessarily involves ‘dislocation’. This paper reviews three influential typologies for the translation of culturally specific items: Ivir (1987), Newmark (1981, 1988), and Hervey and Higgins (1992), referring also to Venuti (1995). It suggests a number of dichotomies for understanding these typologies and the translation of culturally specific items: 1 Source Culture-/Source Language-oriented (domesticating) vs. Target Culture-/Target Language-oriented (foreignising); 2 nonlexicalised/ ungrammatical vs. lexicalised/grammatical; 3 semantically systematic vs. semantically anomalous; 4 synonymy-oriented vs. nonsynonymy oriented; 5 situationally equivalent vs. culturally analogous; 6 lexical vs. structural. As an aid to understanding these typologies, the paper provides a visual ‘grid’, siting the various procedures proposed by each of the four typologies.

Keywords: translation, culture, Arabic, domestication, foreignisation

Introduction This paper considers the translation of culturally specific items, as delimited by the following extreme procedures: i. In the translated text (Target Text) artificially including Source Culture-specific aspects of the original text (Source Text), by

44

Chapter Three

extending the margins of the Target Language and Target Culture through ‘cultural borrowing’; or: ii. In the Target Text artificially presenting elements in the Source Text which are Source Culture-specific as if they were central elements of the Target Culture through ‘cultural transplantation’. The proposals of Ivir (1987), Newmark (1981, 1988), and Hervey and Higgins (1992) are considered in detail because these are the best known and arguably the most coherently worked out sets of proposals in the literature. Venuti (1995), although less specific, will also be discussed, because of the important general orientation provided by his distinction between foreignisation and domestication. The approaches taken in these proposals are summarised in figure 4.1. The following discussion will make extensive reference to that figure, and the various columns it contains.

Source Culture-/Source Language-oriented vs. Target Culture-/Target Language-oriented, and Foreignising vs. Domesticating The most general distinction in respect of culture-specific items is whether the translation is oriented towards the Source Culture and, by extension, Source Language, or the Target Culture, and by extension Target Language. I assume that orientation towards the Source Culture implies also orientation towards the Source Language, and that orientation towards the Target Culture implies also orientation towards the Target Language. I also identify Source Culture-/Source Language-oriented with foreignising and Target Culture-/Target Language-oriented with domesticating (Venuti, 1995), domesticating translation procedures being those given in columns 1, 2 and 3 in figure 4.1, while foreignising translation procedures are given in columns 5, 6 and 7. Culture-neutral is used in figure 4.1, column 4 to refer to a translation which is neither foreignising nor domesticating, but is equally appropriate to both the Source Culture and Target Culture. The boundaries between foreignising and culture-neutral, and between culture-neutral and domesticating are ‘fuzzy’: we cannot always be sure whether a particular element of translation is better defined as foreignising or culture-neutral, or culture-neutral or domesticating. Even within a single language cultural identity is complex: is curry an Indian dish because that is where it originated, or is it now also a British one because Indian restaurants and take-aways are extremely popular in Britain, and millions of people in Britain have curry for tea every night?

Procedures for Translating Culturally Specific Items

Key: Ivir = Ivir (1987); Newmark = Newmark (1981, 1988); H+H = Hervey and Higgins (1992) Figure 4.1 Procedures for translating culturally specific items.

45

46

Chapter Three

Non-lexicalised/Ungrammatical vs. Lexicalised/Grammatical ‘Non-lexicalised’ translation procedures are shown in figure 4.1, column 1. ‘Non-lexicalised’ means that the word in question is not a regular part of the language. By definition, non-lexicalised words are not found in dictionaries. Mizmar, used for example as the English translation of the Arabic έΎϣΰϣ, is an example of a non-lexicalised word. ‘Ungrammatical’ means that the form in question does not conform to the standard grammar of the language. A translation of ϦϴΘΑήο ϲϨΑήο as ‘he beat me two beatings’ is ungrammatical: the adverbial use of a noun phrase cognate to the verb is not part of the grammar of English. Nonlexicalised words are sometimes referred to as nonce-words, while ungrammatical forms can be referred to as nonce-formations (cf. Crystal 2003). The boundaries between what is lexicalised and what is not are not always clear. ‘Islam’ is a well-established lexicalised word in English. ‘Sharia’ (also ‘sheria’) (i.e., Δόϳήη) is given in Collins English Dictionary, but is likely to be unknown to many non-Muslims in Britain. While we might regard ‘sharia’ as lexicalised in a general sense, for those English speakers who do not know it, we may say that it is non-lexicalised.

Semantically Systematic vs. Semantically Anomalous ‘Semantically systematic’ translation procedures are shown in figure 4.1, columns 3-7. ‘Semantically systematic’ means ‘a standard part of the semantic system of the language’. For example, the meanings of ‘fox’ as (1) any canine mammal of the genre Vulpes and related genera, and (2) a person who is cunning and sly, are semantically systematic in English. The meanings of ‘round the bend’ as (1) ‘around the corner’ and (2) ‘mad’, are also semantically systematic in English. In both these cases all the meanings given can be found in a reliable dictionary. ‘Semantically anomalous’ translation procedures are shown in figure 4.1, columns 1-2. ‘Semantically anomalous’ means ‘not part of the semantic system of the language’. The use of ‘aardvark’ to mean ‘an incompetent person’ is semantically anomalous. This is reflected in the fact that ‘aardvark’ is not given in the sense ‘an incompetent person’ in reliable English dictionaries. Similarly, ‘beyond the turning’ in the sense ‘mad’ is semantically anomalous, as reflected in the fact that ‘beyond the turning’ is not glossed as ‘mad’ in reliable English dictionaries. Nonlexicalised words (nonce-words) are by definition semantically anomalous.

Procedures for Translating Culturally Specific Items

47

Since mizmar is not lexicalised in English (not part of the vocabulary of the language), it cannot have a proper (systematic, fixed) meaning. Ungrammatical forms are similarly by definition semantically anomalous. Since ‘he beat me two beatings’ is not part of the grammar of English, it cannot, similarly, have a proper (fixed, systematic) meaning. The fact that forms are non-systematic does not necessarily mean that they cannot be understood (or at least partially understood). In a phrase ‘he blew a beautiful long, single note on the mizmar’, it is fairly clear that the mizmar must be a form of wind instrument. Similarly, it is likely that a native English speaker would understand the phrase ‘he beat me two beatings’ even if they recognise that it is not English.

Synonymy-oriented vs. Problem-avoidance Oriented vs. Non-synonymy Oriented ‘Synonymy-oriented’ translation procedures are shown in figure 4.1, columns 1-4. ‘Synonymy-oriented’ is not used here to mean ‘synonymous’. Rather, it means that the translation is likely to be close to synonymous—even if it is more specifically hyponymous (particularising), hyperonymous (generalising), or semantically overlapping (Dickins, Hervey and Higgins 2002, pp. 54-59), and that it can be reasonably analysed in relation to the notion of synonymy. Non-lexicalised words can be regarded as synonymous with their Source Text forms. Thus the nonlexicalised and semantically anomalous mizmar is—if we are to say that it has any sense at all in English—best regarded as synonymous with the Arabic έΎϣΰϣ. Similarly, ungrammatical forms, such as ‘two beatings’ (in ‘he beat me two beatings’) are best regarded as ‘structurally synonymous’ with their Source Text originals—i.e. ‘two beatings’ here is to be regarded as having an adverbial sense in English. In figure 4.1, column 5, I have identified omission as a cultural translation procedure with problemavoidance: by not attempting to find any equivalent for the Source Text word or phrase, the problem of what an appropriate equivalence might be is avoided. Non-synonymy oriented translation procedures are given in figure 4.1, columns 6 and 7. Non-synonymy oriented translation procedures are those in which the issue of synonymy is not of focal importance. Non-synonymy oriented translations are domesticating in that they involve use of specifically Source Culture-oriented uses of language.

48

Chapter Three

Situationally Equivalent vs. Culturally Analogous Figure 4.1 includes two types of non-synonymy oriented translations The first, situational equivalence (column 6), involves cases in which the same situations (or functions) can be identified in both cultures. Thus, people see others off on a journey in both Western and Arabic culture. In Britain, one might say to someone one is seeing off, ‘Have a nice journey’, or ‘Have a safe journey’, or even ‘All the best’. In Sudan, the standard phrase is ௌ ϚΘϋΩϭ (or ௌ ϙΎϨϋΩϭ). These phrases are situationally (or functionally) equivalent; whether they are nearly synonymous or not nearly synonymous is of secondary importance. The second type of non-synonymy oriented translation, that of cultural analogy (column 7), is where there is no obvious situational equivalent in the Target Text Culture: that is to say, the particular situation—or feature—in question is part of the Source Culture, but not part of the Target Culture. Culture-specific literary allusions often give rise to this kind of case. Thus ϰϠϴϟϭ βϴϗ as an ironic description of two young lovers (cf. Dickins, Hervey and Higgins 2002, p. 32) involves a literary allusion which is specific to Arab (and more generally Middle Eastern) culture. Precisely the same ‘situation’ (i.e., characters) does not occur in Western culture. However, in English literature, and therefore English-language culture, Romeo and Juliet—as doomed lovers—occupy an analogous situation to that of βϴϗ and ϰϠϴϟ in Arab culture. ϰϠϴϟϭ βϴϗ may therefore, in some circumstances, be replaced by Target Text ‘Romeo and Juliet’ by a process of cultural analogy. There are cases which fall somewhere between situational equivalence and cultural analogy. When someone has had their hair cut, it is customary in many Arabic countries to say ˱ ΎϤϴόϧ meaning ‘with comfort/ease’, to which the standard reply is ϚϴϠϋ ௌ Ϣόϧ΃ (with some variants) ‘may God grant you comfort/ease’. English has, of course, the cultural situation of haircutting—there is no need here to search for a cultural analogy. What it lacks, however, is any standard phrase which is uttered when someone has their hair cut: there is no real situational equivalent.

Lexical vs. Structural (Morphotactic or Syntactic) Row A and row B (columns 1-3 only) distinguish between lexical and structural translation procedures. In the case of foreignising translations not involving omission, the foreignising element may be lexical (row A, columns 1-3), i.e., a feature of the words used (considered as single units). Alternatively, it may be structural (row B, columns 1-3), i.e., a feature of

Procedures for Translating Culturally Specific Items

49

the way in which words are put together from individual morphemes (morphotactic) or the way in which words themselves join together to form larger phrases (syntactic)—or both. What Hervey and Higgins call cultural borrowing (see below) is normally a case of a monomorphemic word, i.e., a word which consists of only one morpheme: as such it is lexical rather than structural. For example, in Yemen the word ΏΎ͉ΑΩ˴ is used for a particular type of minibus (normally a Toyota mini-van). In Arabic, this is likely to be analysed as consisting of two morphemes: the root Ώ Ώ Ω and the pattern ϝΎ͉ό˴ϓ. If, however, we use the cultural borrowing dabab to translate the Arabic ΏΎ͉ΑΩ˴ , the form in English consists of a single morpheme: the grammar of English does not allow us to identify separate root and pattern morphemes here. In the case of ungrammatical calque the foreignising element is structural. That is to say, it is either morphotactic, or syntactic. ‘He beat me two beatings’ consists of standard English words: the overall form, however, is structurally (syntactically) foreignising.

Plotting Ivir’s, Newmark’s, and Hervey and Higgins’ Procedures In the following sub-sections, I will firstly provide proposed general descriptions—some of which have also been used, or are usable, as terms—of the translation procedures defined by figure 4.1. These descriptions are presented in unboxed text in figure 4.1. Following that I will consider the specific procedures proposed in Ivir (1987), Newmark (1988), and Hervey and Higgins (2002) presented in boxes, as these are classified according to column and row (for columns 1-3) in figure 4.1. Because some of the procedures of Ivir, Newmark and Hervey and Higgins belong to more than one column and/or row, I will group these authors’ procedures together in common-sense categories, in order to present the information in a manner which is relatively coherent and comprehensible.

Cultural Borrowing Proper and Ungrammatical Calque/Exoticism The column 1, row A translation procedure could be termed cultural borrowing proper. The column 1, row B procedure could be termed ungrammatical calque/exoticism. Dickins, Hervey and Higgins (2002) specifically confine cultural borrowing to non-lexicalised lexical (nonstructural) forms—i.e. column 1, row A. Thus, dabab—see section Lexical

50

Chapter Three

vs. structural (morphotactic or syntactic), above—is a cultural borrowing, but intifada, as a translation of ΔοΎϔΘϧ΍, is no longer a cultural borrowing, on the grounds that intifada has now become a regular part of the English language (cf. Dickins, Hervey and Higgins 2002, p. 34). Ivir seems to define as a borrowing any conspicuously foreign-derived word. This may be non-lexicalised, e.g., dabab as a translation of the Yemeni ΏΎ͉ΑΩ˴ ., in which case it belongs to column 1. Or, it may be lexicalised, e.g., intifada, in which cases it belongs in column 3. In figure 4.1, I have connected the two Ivir ‘borrowing’ boxes in column 1 and column 3 with a double-headed arrow, to indicate that they constitute, for Ivir, a single procedure. Since Ivir is talking about borrowing without structural complexity (both dabab and intifada consist of a single morpheme in English), Ivir’s borrowing belongs entirely in row A. Newmark defines transference as “the process of transferring a Source Language word to a Target Language text as a translation procedure. It [...] includes transliteration, which relates to the conversion of different alphabets: e.g., Russian (Cyrillic), Greek, Arabic, Chinese, etc. into English. The word then becomes a ‘loan word’” (Newmark 1988, p. 81). Newmark includes within the procedure of transference both nonlexicalised terms (e.g., dabab) and lexicalised terms (e.g., intifida). Like Ivir’s ‘borrowing’, Newmark’s ‘transference and naturalisation’ thus belongs in both column A and column C (as with Ivir’s ‘borrowing’, I have connected Newmark’s ‘transference and naturalisation’ boxes with a double-headed arrow, to show that for Newmark these two boxes constitute a single procedure). Newmark’s transference is a simple adoption of a word (or phrase) without any adaptation to the Target Language. Transference is by definition lexical—there is no internal structuring (whether morphotactic or syntactic) in the Target Language form—and thus belongs to row A. Naturalisation “succeeds transference and adapts the Source Language word first to the normal pronunciation, then to the normal morphology (word forms) of the Target Language” (Newmark 1988, p.82). Where naturalisation involves only phonological adaptation, it remains lexical: intifada, when pronounced by the average English speaker may sound English (almost rhyming with, for example, ‘winter larder’). However, as it is morphologically simple in English, it is lexical, rather than morphotactic. One example which Newmark gives of naturalisation is French thatchérisme, from English ‘Thatcherism’ (the political philosophy associated with Margaret Thatcher). Here the root element thatchér has been somewhat adapted to French pronunciation and spelling—thatchér as

Procedures for Translating Culturally Specific Items

51

opposed to English ‘Thatcher’, the suffix has the standard French form isme (English ‘-ism’), and the word is spelt with an initial lower-case letter ‘t’, rather than the English upper-case ‘T’. In the case of thatchérisme, the French form is morphologically complex—consisting of the morphemes thatchér and -isme (assuming that -isme is morphologically simple). Thatchérisme is thus structural (morphotactic) and belongs to row B. By ‘lexical creation’ Ivir seems to mean non-lexicalised words, newly invented by the translator out of existing morphological elements in the Target Language. As such Ivir’s ‘lexical creation’ belongs in column 1 (non-lexicalised), row B (structural—morphotactic). Ivir may also mean to include words involving more than one morpheme in the Target Language which have become systematic in the Target Language, but are still perceived as neologisms—in which case Ivir’s ‘lexical creation’ box should be extended to include also column 3, row B. Under ‘calque’, Hervey and Higgins include forms which are ungrammatical and semantically anomalous, such as ‘it increased the clay moistness’ for ΔϠΑ Ϧϴτϟ΍ Ω΍ί (Dickins, Hervey and Higgins 2002, p. 31). This is ungrammatical, because the adverbial use of ‘moistness’ is not a grammatical feature of English, and semantically anomalous, because it cannot (as an ungrammatical form) have a systematic meaning. These cases belong to column 1 (non-lexicalised/ungrammatical), row B (structural—syntactic). Hervey and Higgins also use the term ‘calque’ to describe forms which are grammatically systematic, but semantically anomalous. An example would be ‘it increased the clay’s moistness’ (in which the ungrammaticality of ‘it increased the clay moistness’ has been eliminated) as a translation of ΔϠΑ Ϧϴτϟ΍ Ω΍ί. Although ‘it increased the clay’s moistness’ is grammatical, it remains semantically anomalous if it is used to express the sense ‘it made matters worse’: this is not a standard meaning of this phrase in English. These cases belong in column 2 (lexicalised/grammatical, but semantically anomalous), row B (structural syntactic). Hervey and Higgins state that “a Target Text marked by exoticism is one which constantly uses grammatical and cultural features imported from the Source Text with minimal adaptation” (Hervey and Higgins 2002, p. 34). An example given (in slightly longer form) in both Thinking French translation (ibid.) and Thinking Arabic translation (Dickins, Hervey and Higgins 2002, pp. 139-140) is the following from the Maqamat of Al-Hariri: ˴ ˵ ˸μΨ˴ η˴ .ω ˶ ή˴ ό˶ ϟ΍ ˴Ϧϣ˶ Ζ ˶ ˸έά˴ ϟ΍ ͊Ϯ˵ϠΧ˵ ϲ˶ϨϴϬ˶ Ϡ˵˸ ϳ . ΔԺ ˴ρϮ˵Βϐ˸ ϣ˴ ΓԺ Ϊ˴ Ο˶ ϭ˴ . ΔԺ ˴ρϮ˵Α ˸ήϣ˴ ΩԺ ˸ή˵Ο ϭΫ˵ Ύϧ΃ϭ˴ . Δ˶ ˴ρϮϐ˵ ϟ΍ ϰϟ· ϕ΍ ˵ ˸ή .ω π ϟ΍ ϝ Ϯ ϔ Σ ϲ ˵ ˵ ˴ ˶Ϩϴϫ˶ Ω˴ ΰ˸ ˴ϳϭ˴ ˶

52

Chapter Three I went from ‘Irák to Damascus with its green water-courses, in the day when I had troops of fine-bred horses and was the owner of coveted wealth and resources, free to divert myself, as I chose, and flown with the pride of him whose fullness overflows.

This Target Text goes beyond the mirroring of grammatical and cultural features—at least if cultural features are defined in a narrow sense—to include replication of prosodic features (rhythm and rhyme) of the Source Text. If we include these additional features as elements of exoticism, the account given of exoticism in figure 4.1 is only partial (since it makes no reference to non-grammatical or non-semantic features). In this respect we can regard ‘exoticism’ as a hyperonym of ‘calque’. The second feature of exoticism which is suggested by Hervey and Higgins’ phrase “constantly uses” is that exoticism is a general orientation throughout a text, whereas calque is “a momentary foreignness” (Hervey and Higgins 2002, p. 34). This distinction is, again, not specifically represented in figure 4.1, which focuses on individual occurrences rather than global Target Text orientations.

Semantic Extension Mirroring Source Language Usage, and Grammatical, but Semantically Anomalous Calque/Exoticism Involving Semantic extension The column 2, row A translation procedure can be described as semantic extension mirroring Source Language usage (‘literal’ lexical equivalent). The column 2, row B translation procedure can be described as grammatical, but semantically anomalous calque/exoticism involving semantic extension (‘literal’ translation of phrase). Hervey and Higgins’ calque and exoticism has been described above. As noted there, cases of calque which are semantically anomalous but grammatical belong in column 2, row B. Ivir’s ‘literal translation’ overlaps with Hervey and Higgins’ calque (or calque/exoticism), and covers both grammatical but semantically anomalous phrases such as ‘it increased the clay’s moistness’ and single words, e.g., the translation of Arabic Δ͉Ϩγ˵ referring to the norms of the Islamic community, by the original basic (literal) meaning of Δ͉Ϩγ˵ ‘path’. Regardless of whether the element in question is a word or a phrase the operative principles are these: 1. The Source Text element (word or phrase) has more than one meaning (or sense), i.e., it is polysemous.

Procedures for Translating Culturally Specific Items

53

2. One of the Source Text element’s senses is basic, while the other relevant sense is secondary. Typically the secondary sense is likely to be perceived as metaphorical, but it may be figurative in some other way, e.g., metonymical. It may even not stand in an unambiguous figurative relationship to the primary sense. Crucially, however, the secondary sense must be clearly conceptually secondary to the primary one. 3. The Target Text element must have the same primary sense as the Source Text element. 4. The Target Text element must not have the same secondary sense as the Source Text element. Consider the English phrase ‘go up the wall’ in relation to a literal Arabic translation έ΍ΪΠϟ΍ Ϊόλ. 1. English ‘go up the wall’ fulfils condition 1: it is polysemous, meaning ‘i. ‘climb the vertical partition (etc.)’, and ii. ‘get very angry’. 2. The Source Text sense ‘climb the vertical partition’ is conceptually primary. The idiomatic sense ‘get very angry’ is perceived as metaphorical. 3. The Target Text έ΍ΪΠϟ΍ Ϊόλ has the same primary sense as ‘go up the wall’. 4. The Target Text έ΍ΪΠϟ΍ Ϊόλ does not have the same secondary sense as the Source Text ‘go up the wall’. (Ϊόλ έ΍ΪΠϟ΍ does not standardly mean ‘get very angry’ in Arabic.) Ivir’s ‘literal translation’ belongs to column 2 in figure 4.1 (semantically anomalous, in that the meaning assigned to the word or phrase is not a meaning which that word or phrase standardly has in the Target Language, but lexicalised/grammatical, in that the word or phrase is a regular part of the lexis/grammar of the Target Language). Where Ivir’s ‘literal translation’ involves only a single word consisting of a single morpheme (or, by extension, where the morphological structure— morphotactics—of this word is not important in translation terms) this is a lexical form (row A). Where Ivir’s ‘literal translation’ involves morphotactic or syntactic considerations, this is a structural form (column 2, row B), έ΍ΪΠϟ΍ Ϊόλ in Arabic, if used in the sense ‘get very angry’ (‘go up the wall’) being an example. I have accordingly shown Ivir’s ‘literal translation’ procedure straddling rows A and B (column 2) in figure 4.1. What Newmark means by ‘literal translation’ seems to be the same as what Ivir means by ‘literal translation’, and therefore also straddles rows A and B in column 2 in figure 4.1.

54

Chapter Three

Lexicalised Cultural Borrowing, and Grammatically and Semantically Systematic Calque/Exoticism The column 3, row A translation procedure could be termed lexicalised cultural borrowing. The column 3, row B translation procedure can be described as grammatically and semantically systematic calque/exoticism. The reasons Ivir’s borrowing belongs in both column 1 and column 3 (row A), and why Newmark’s transference and naturalisation belongs in both column 1 and column 3 (row A and row B) have been discussed above (in the section Cultural borrowing proper and ungrammatical calque/ exoticism). Newmark defines ‘through-translation’ as “the literal translation of common collocations, names of organisations, the components of compounds” (Newmark 1988, p. 84). However, unlike Newmark’s ‘literal translation’ (see discussion in section Cultural borrowing proper and ungrammatical calque/exoticism above), which is semantically anomalous (column 2) and may be lexical or structural (rows A or B), his ‘throughtranslation’ is semantically systematic (as well as foreignising) (column 3), and structural (morphotactic or syntactic) (row B). Examples given by Newmark include ‘superman’ from German Übermensch (über meaning ‘above, over’, Mensch meaning ‘man, human being’). Newmark’s procedure of ‘through-translation’ is similar to Hervey and Higgins’ calque/exoticism, and Newmark himself notes that literal translation is also “known as calque or loan translation” (Newmark 1988, p. 84). However, whereas Hervey and Higgins’ calque (see section Cultural borrowing proper and ungrammatical calque/exoticism above) is semantically anomalous, Newmark’s through-translation is, as noted, semantically systematic.

Culture-neutral Word/Phrase In columns 4-7, we move away from translation procedures which are Source Culture/Source Text oriented. The distinction between lexical (row A) and structural (row B), which was important for considering how the elements of the Target Language-form relates to those of the Target Language-form for procedures in columns 1-3, no longer obtains, and is thus not made in figure 4.1 for columns 4-7 This translation procedure could be termed culture-neutral word/phrase. ‘Descriptive equivalent’ in Newmark seems to mean the same as ‘defining’ in Ivir (below). This can be regarded as a culture-neutral procedure. It involves a fairly precise description of what is meant by the

Procedures for Translating Culturally Specific Items

55

Source Culture element. However, it achieves this through the use of words and phrases which are generally understood in the Target Culture. Newmark’s ‘descriptive equivalent’ belongs in column 4. Among the examples which Newmark gives of descriptive equivalence is “the Japanese aristocracy from the eleventh to the nineteenth century” for Samurai. ‘Functional equivalent’ in Newmark is somewhat more difficult to understand. Examples given by Newmark (1988, p. 83) are: baccalauréat ‘French secondary school leaving exam’, and Sejm ‘Polish parliament’. Newmark says of functional equivalence that “[t]his procedure occupies the middle, sometimes the universal, area between the Source Language language or culture and the Target Language language or culture” (Newmark 1988, p. 83). He goes on, “[i]n translation, description sometimes has to be weighed against function. Thus, for machete, the description is a ‘Latin American broad, heavy instrument’, the function is ‘cutting or aggression’. Description and function are combined in ‘knife’. Samurai is described as ‘the Japanese aristocracy from the eleventh to the nineteenth century’; its function was ‘to provide officers and administrators’” (Newmark 1988, pp. 83-84). ‘Descriptive equivalent’ in Newmark seems to answer the question ‘What is it?’, while ‘functional equivalent’ seems to answer the question ‘What does it do?’. I have analysed both as culture-neutral, and as synonymy-oriented (column 4). ‘Functional equivalence’ might appear to be less synonymy-oriented than ‘descriptive equivalence’. In the case of tools (and similar) made by human beings for a purpose (or function), however, that purpose seems to be part of the definition. For example, a gimlet (a hand tool for boring small holes in wood) may look exactly like a small screwdriver: it is only because the intention is that this tool should bore holes in wood, rather than putting in screws into wood (or taking them out) that we classify it as a gimlet and not as a screwdriver. Given that function can be an essential part of the definition of an object, I have placed ‘functional equivalent’ directly next to (below) ‘descriptive equivalent’. However, it might also be possible to interpret ‘functional equivalent’ in another way—as what is appropriate (‘functionally appropriate’) in a given situation; e.g., what one says when bidding farewell to a friend, or on finishing a meal. In this case, Newmark’s ‘functional equivalence’ could be regarded as identical to Hervey and Higgins’ ‘communicative translation’ (column 6). To indicate this possibility, I have put a single-headed arrow from Newmark’s ‘functional equivalent’ in column 4 to column 6.

56

Chapter Three

‘Defining’, in Ivir, typically involves textual expansion (additional words/phrases are used). We may, however, come across situations in which a definition is briefer than the original Source Text usage, in which case we can refer to this as (culture-neutral) contraction. The most extreme form of contraction is omission (section Omission for cultural reasons, below). Together with defining, Ivir mentions the procedure of addition, i.e., when additional information is added in the Target Text which is not in the Source Text. Addition comes very close to definition, and I have included it immediately below ‘definition’ in figure 4.1. In column 4, I have included a vertical double-headed arrow, to show that culture-neutral translation procedures may vary from contraction at one extreme to expansion at the other. ‘Explanation’ in Hervey and Higgins seems to mean the same as ‘defining’ in Ivir and ‘descriptive equivalent’ in Newmark. This procedure frequently occurs together with (cultural) borrowing (column 1), i.e., the foreignism is introduced, and the Target Text subsequently (or perhaps immediately before) makes plain, either directly or in a less explicit way, what the foreignism means.

Omission for Cultural Reasons The column 5 translation procedure could be termed omission for cultural reasons. As noted above (in the section Synonymy-oriented vs. problem-avoidance oriented, vs. non-synonymy oriented), omission involves avoiding the normal problems associated with translating a culturally specific element. It can be regarded as domesticating in that it removes mention of the foreign element in the Target Text. Newmark does not specifically discuss omission as a cultural translation procedure and I have not therefore included Newmark in column 5. He does, of course, recognise the possibility of omission in translation. Dickins, Hervey and Higgins (2002, pp. 23-24) discuss omission as a translation procedure, but stress that it may have a number of different purposes—not all of them to do with culture. I have not therefore included Hervey and Higgins in column 5.

Communicative Translation The column 6 translation procedure could be termed communicative translation. “A communicative translation is produced, when, in a given situation, the Source Text uses an Source Language expression standard for that situation, and the Target Text uses a Target Language expression

Procedures for Translating Culturally Specific Items

57

standard for an equivalent Target Culture situation” (Dickins, Hervey and Higgins 2002, p. 17), public notices, proverbs and conversational clichés providing good examples: ϦϴΧΪΘϟ΍ ωϮϨϤϣ ΪΣ΍ϭ ήΠΤΑ ϦϳέϮϔμϋ Ώήο ΐΟ΍ϭ ϰϠϋ ήϜηϻ

No smoking (public notice) To kill two birds with one stone (Standard Arabic proverb) Don’t mention it (conversational cliché)

Communicative translation does not involve referring to something in the Target Culture which does not exist in the Source Culture. Rather, it involves using a phrase (or possibly a single word) in a context in the Target Text where this phrase (or word) is typically used in the Target Culture, as a translation of a phrase (or word) used in the Source Text which is typically used in this context in the Source Text, and where the meaning (and particularly the denotation) of the Target Text phrase (or word) is clearly different from that of the Source Text phrase (or word). An example given by Hervey and Higgins (1992) is Chinese Source Text (back-translated) ‘How many persons in your family?’ in the context of a greeting routine, translated into an English Target Text as ‘Nice weather for the time of year’. After greeting one another, strangers in China typically ask about one another’s family. In Britain, by contrast it is culturally normal to ask about the weather. Families and weather are aspects of culture (or life) in both China and Britain. The contexts in which these two topics are typically talked about are, however, rather different in the two cultures. Ivir does not have an equivalent of Hervey and Higgins’ communicative translation. It is worth recognising a cline for communicative translation. At one extreme, there may be only one Target Language equivalent for a Source Language word or phrase. For example, in a particular culture (and language), there may be only one thing which it is standardly possible to say in condoling someone about a mutual friend’s death. At the other extreme, however, there may be numerous things one can standardly say in a particular situation in a particular culture (and language). Thus, in seeing a friend off in English, one can standardly say a number of things such as ‘Have a nice / good / pleasant trip / journey’, ‘Look after yourself’, ‘Goodbye’. These are multiple alternative communicative equivalents of what may be only one single possible phrase in a Source Language. The cline between a ‘unique equivalent’ and ‘multiple equivalents’ in communicative translation is recognised in column 6 by a vertical doubleheaded arrow.

58

Chapter Three

Newmark (1981, pp. 36-69) uses the term ‘communicative translation’, but means something much wider than what Hervey and Higgins mean by it. Newmark’s notion of ‘communicative translation’ is thus not directly relevant here, and has not been included in figure 4.1. As noted in the section Culture-neutral word/phrase above, however, Newmark’s functional equivalent—understood in a certain way—could be regarded as the same as Hervey and Higgins’ communicative equivalent. We can regard Chinese ‘How many persons in your family?’ as fulfilling the same function—that of making polite conversation between strangers—as does English ‘Nice weather for the time of year’. The two phrases could, therefore, in this context be said to be functionally equivalent.

Cultural Transplantation The column 7 translation procedure could be termed cultural transplantation (as in Hervey and Higgins 1992). Newmark terms it cultural equivalent. As discussed in the section Situationally equivalent vs. culturally analogous (above), where there is no situational identity, communicative translation is impossible. One may in these cases invoke the notion of cultural analogy. If the same elements are not found in both cultures, the translator may substitute something in the Target Text from the Target Culture which is similar to the element referred to in the Source Text in the Target Culture. Newmark refers to this substituted element as a cultural equivalent. Examples given by Newmark (1988, p. 83) are British ‘cricket’ or American ‘baseball’ (common sports in Britain and America respectively) as translations of French le cyclisme (cycling), which is a very common sport in France, but less so in Britain or America. Ivir’s ‘substitution’ is the same as Newmark’s ‘cultural equivalent’. Hervey and Higgins define cultural transplantation on a large scale as “the wholesale transplanting of the entire setting of the Source Text, resulting in the entire text being rewritten in an indigenous Target Culture setting” (Dickins, Hervey and Higgins, 2002, p. 32). They give as an example of wholesale cultural transplantation the remaking of the Japanese film The Seven Samurai as the Hollywood film The Magnificent Seven, but point out that in translation a much more likely procedure is small-scale cultural transplantation, e.g., the replacement of Source Text ϰϠϴϟϭ βϴϗ by Target Text ‘Romeo and Juliet’ (cf. section Situationally equivalent vs. culturally analogous, above). It is this small-scale cultural transplantation which most closely corresponds to what Newmark means by ‘cultural equivalent’ and Ivir by ‘substitution’.

Procedures for Translating Culturally Specific Items

59

Conclusion I have argued that the translation of culturally specific items involves various procedures, ranging from extension of the margins of the Target Language and Target Culture at one extreme, to artificially presenting elements in the Source Text which are Source Culture-specific as if they were central elements of the Target Culture at the other. I have established a conceptual ‘grid’ (figure 4.1) which compares the procedures recognised by Ivir, Newmark, and Hervey and Higgins. Beyond this, however, the current account also provides a synthesis of previous approaches, by placing these procedures within a unified conceptual framework. It thus in fact presents a new model of procedures for translating culturally specific items—one which has more categories, and whose categories are, I believe, more coherently defined with respect to one another than are those of previous accounts.

Bibliography Crystal, D. 2003. A Dictionary of Linguistics and Phonetics. Oxford: Blackwell. Dickins, J. 2005. Two models for metaphor translation. In Target 17: 2, pp. 227-273. Dickins, J., S.G.J. Hervey and I. Higgins. 2002. Thinking Arabic Translation. London and New York: Routledge. Hervey, S.G.J. and I. Higgins. 2002. Thinking French Translation. London and New York: Routledge. Hervey, S.G.J. and I. Higgins. 1992. Thinking Translation: A Course in Translation Method: French to English. London and New York: Routledge. Ivir, V. 1987. Procedures and strategies for the translation of culture. In Toury, G. (ed). Translation Across Cultures, pp.35-46. New Delhi: Bahri. Newmark, P. 1981. Approaches to Translation. Oxford: Pergamon. —. 1988. A Textbook of Translation. New York: Prentice Hall International. Munday, J. 2002. Introducing Translation Studies. London and New York: Routledge. St John, J. 1999. Translation of ΞδϔϨΒϟ΍ ϞϘΣ and ˯ΎϤϟ΍ϭ έΎϨϟ΍ by ήϣΎΗ Ύϳήϛί (1973. In ϖ΋΍ήΤϟ΍ ϖθϣΩ. Damascus: έ΍ϮϧϷ΍ έ΍Ω). BA translation project: University of Durham.

60

Chapter Three

Venuti, L. 1995. The Translator’s Invisibility. London and New York: Routledge. —. 1998. The Scandals of Translation: Towards An Ethics of Difference. London and New York: Routledge.

CHAPTER FOUR PROVERB TRANSLATION: FLUENCY OR HEGEMONY? AN ARGUMENT FOR SEMANTIC TRANSLATION ABDUL GABBAR AL-SHARAFI

Abstract Based on the common conception that translation is primarily a cultural act, this paper aims to investigate the cultural implications of applying two different translation approaches: a semantic approach and a communicative approach to the translation of English proverbs into Arabic. The former is a source-text and source-culture preserving translation strategy, while the latter is a target-oriented translation strategy which caters to target readers' expectations in terms of naturalness, fluency and readability. This is illustrated through two different types of translations, discussed in terms of their cultural implications. This paper argues that a semantic approach is more appropriate because it reflects the essence of translation which is cultural exchange by means of source culture preservation and target culture enrichment.

Keywords: culture, semantic translation, communicative translation, naturalness, hegemony, domestication, foreignisation, semiotics

Introduction Due to their rich connotations, special form and perceived cultural value, translating proverbs is usually considered to be difficult. Nonetheless, proverbs have been translated using two main approaches. One of these two main approaches is the semantic approach which, as will be explained later, aims to preserve the source culture and take the target

62

Chapter Four

reader to the source culture. The second approach is the communicative approach which, as will be explained, aims to naturalise the translation to meet target readers' expectations and thereby keeps the target readers where they are and brings the foreign text naturalised and domesticated for them. This paper argues against the second approach which has evolved into a modern preference for ‘fluent translation’ that has become prevalent particularly with regard to proverb translation. According to this ‘fluency view’, a translated text, whether prose or poetry, fiction or nonfiction, is judged acceptable by most publishers, reviewers and readers when it reads fluently, when the absence of any linguistic or stylistic peculiarities makes it seem transparent, giving the appearance that it reflects the foreign writer's personality or intention or the essential meaning of the foreign text–the appearance, in other words, that the translation is not in fact a translation, but the ‘original’ (Venuti, 1995, p. 1).

This fluency, it is claimed, is motivated by a search for clarity, smoothness and naturalness of style. The underlying motive might even be discomfort at knowing the other and a tendency to assimilate the other, rather than appreciate and understand the difference. This rather disconcerting trait seems to be particularly manifest in proverb translation where the most extreme form of domestication is usually practiced by those who call themselves ‘successful’ and who measure their ‘success’ by the degree of departure from the source text's semantic form by replacing the foreign proverb with a local situational equivalent. The same fluency preference is also evident in proverb translation from Arabic into English. For example, El-Yasin and Al-Shehabat (2005) argue that "optimal [proverb] translation is that in which the translator can achieve optimal equivalence; that is to say, it is the translation which creates in the target-language audience the same effect that the source language (SL) text creates on its audience" (p. 163, emphasis mine). By ‘effect’, the authors mean the desired response expected from the target readers. As-Safi and ash-Sharifi (1997) set naturalness as their main consideration for literary translation. A key feature of naturalness, according to them, is ‘idiomaticity’, which "implies that wherever established formulas—collocations, idioms and proverbs—exist in a language, they should be used rather than free word combinations'' (p. 71). Abu Ssaydeh supports a domesticating approach to the translation of English idioms into Arabic when he says that “the seasoned translator will give priority in idiom translation to those that are identical in TL” (2004,

Proverb Translation: Fluency or Hegemony?

63

p. 118). This paper's main argument is that although translation strategies such as domestication, localisation, adaptation and cultural transplantation might be appropriate for translating instrumental or commercial texts (i.e. texts of a day-to-day practical nature), such strategies are inappropriate for scholarly, artistic or documentary texts like proverbs which are texts of a special form, are of a literary nature and are culturally valued. Two norms of semantic translation are identified and it is argued that they are systematically marginalised in translating proverbs for the sake of a dominant trend that seeks fluency in translation. These norms are: (1) source culture preservation (2) target culture enrichment Norm (1) essentially means that proverb translation should aim to preserve source culture references because proverbs usually carry with them a rich source-cultural heritage. Norm (2) implies that proverb translation should aim to enrich the target culture because, by their nature, proverbs tend to educate. A number of proverb definitions are reviewed here and a semiotic definition is provided and related to a triadic perspective of translation, thus treating both, proverbs and their translation, as signs. Then, semantic translation is defined and the merits of this approach are discussed using, as illustrations, two English proverbs semantically translated into Arabic. Thereafter, communicative translation is also defined, its disadvantages discussed, and communicative translations of the same two English proverbs examined. The examples are taken from the Oxford Dictionary of English Proverbs (1970), their semantic translations into Arabic are taken from Attia (2004) and their communicative translations are taken from Ba'labki (2001).

Towards a Semiotic Definition of Proverbs In her introduction to the third edition of The Oxford Dictionary of English Proverbs (ODEP), Joanna Wilson (1970) argues that defining a proverb is a rather difficult task. She points out that "...men who have worked for years on the subject have hardly bettered the dictionary definition" (p. vii). She then cites G. L. Apperson's definition which calls proverbs "a crystallized summary of popular wisdom or fancy" (p. vii). The notion of wisdom in Apperson's definition reflects the cultural importance of proverbs as sources of insight and astuteness: hence their high cultural and scholarly value. Norrick (1985) formulates what he calls an ethnographic definition of the proverb as "a traditional, conversational, didactic genre with general meaning, a potential free conversational turn, preferably with figurative meaning" (1985, p. 78).

64

Chapter Four

Looking at proverbs from a semiotic perspective helps us to perceive them as signs and to establish a relationship between proverbs and translation as signifying systems. According to Charles Sanders Peirce, "a sign is something by knowing which we know something more" (Peirce cited in Eco, 1984, p. 2). For something to qualify as a sign "it must have a physical form, it must refer to something other than itself, and it must be recognized as doing this by other users of the sign system" (Turner, 1992, p. 17). The three criteria proposed by Turner correspond to some proverb features suggested by Norrick. The physical form corresponds to the conversational nature of proverbs in that they must be embodied in some kind of physical material which is sound in the form of waves that can be heard, or writing in the form of physical graphological symbols which can be seen. The fact that proverbs must have a general meaning corresponds to Turner's second criterion above that signs must refer to something other than themselves. The fact that proverbs are traditional relates to Turner's third criterion that this referential or general meaning must be recognised by other users in the sign system, i.e. the speech community. Peirce's semiotics perceives the concept of sign as comprising three facets: the representamen (R), the interpretant (I), and the object (O). The first, the representamen, is the signifier and can be in a verbal or nonverbal form. This can generally be called the formal facet of signification, i.e. the expression itself, as in the word ‘next’. The second facet of signification is the interpretant or semantic level of signification which essentially gives the denotative dimension of meaning, the ‘sense’ in Frege's (1966) account; here it is the denotative or dictionary meaning of the word 'next'. The third facet is the ‘object’, the pragmatic facet which is ‘reference’ in Frege's account. This can be called the contextual or situational dimension of meaning. It deals essentially with connotative, associative and other forms of extra-linguistic associations. In the example given, this facet will be the pragmatic meaning of the word ‘next’ said in the context of a queue in a bank or a shop. Here the word ‘next’ acquires an extra meaning from context and it is interpreted as a request for the next person in the queue to come forward to be served. The three facets of meaning then are formal, semantic and pragmatic respectively. This conception helps the author to develop the three triadic models in figure 9.1 and relates the first model of Sign to another triadic model of representation of Proverb matching the three facets. The first facet of representation is the formal one (F) which is the level of proverbial expression. The second is the internal or denotative meaning of the proverb, which is termed here the semantic facet (S). The third is the pragmatic meaning (P) or the second meaning of the proverb which is

Proverb Translation: Fluency or Hegemony?

65

always inherent in it and is expressed in Norrick's feature matrix as didactic and figurative. A proverb in this conception has three facets of signification: the formal (F), the semantic (S), and the pragmatic (P). This semiotic perspective also helps the author to relate these two triadic models of Sign and Proverb to another triadic model of Translation signification. Formal translation (F) represents a concern with translating forms, which is usually the focus of form-focused texts. Semantic translation (S) represents a concern with translating content-focused texts. It is the type of translation seen to give justice to source culture content. Communicative translation (C) represents a concern with fluency and provides the target text reader with (to use Martin Luther's metaphor cited in Reiss, 2000) a field that is clean of all boulders and clods that can be plowed smoothly and easily. The arrows in the diagram show the trajectory of interpretation. One first hears or reads the expression, the R of Sign, the F of Proverb, and the F of Translation, at the semiotic level in which the R and F can be rendered into another language. Then one decodes the I of SIGN, the S of Proverb, and the S of Translation at the semiotic level in which the I and S can be rendered semantically into another language. Next, one interprets the O of Sign and the P of Proverb, and the C of Translation as the level where the intended meaning can be communicatively rendered into another language. Figure 9.1 illustrates the three triadic representations of sign, proverb and translation.

T

S

O

R

SIGN

F

S

P

PROVERB

F

C

TRANSLATION

Figure 9.1: Semiotic representation of the three concepts of sign, proverb and translation

66

Chapter Four

Semantic Translation, Cultural Exchange and Cultural Responsibility This section provides a definition of semantic translation. Newmark (1981) defines this as the type of translation ''where the translator attempts, within the bare syntactic and semantic constraints of the TL, to reproduce the precise contextual meaning of the author" (p. 22). Munday (2008) provides a summary of Newmark's (1981) account of this mode by stating that, in semantic translation, "the focus is on the thought processes of the transmitter as an individual...it remains within the SL culture,...is appropriate for serious literature, autobiography, personal effusion" (p. 45). It is this focus on the thought processes of the transmitter, whether as an individual or as a culture, that should be preserved in the translation process of such serious literature as proverbs which constitute the cultural reservoir of a speech community. It is believed that the ethical aim of translation, as expressed by Berman (2004), is "receiving the foreign as foreign" (p. 277). The significance of translation in von Humboldt's words is clearly seen when he maintains that translation "is one of the most important functions in a literature, partly because it leads people who have no knowledge of other languages to those forms of humanity and the arts which constitute the peculiar significance of every nation, and which they would never experience otherwise, but also partly and even more significantly because they expand the value and expressiveness of their own language" (cited in Reiss, 2000, p. 88). Humboldt reflects the two norms proposed in this paper, namely, source culture preservation and target culture enrichment. These two norms are in fact the essence of the translating act, especially when this act deals with literary or scholarly texts as in the case of proverbs. The universality of proverbial materials preserved in almost every culture is not contested here; rather it is argued that this universal material (usually couched in slogans of wisdom) is always expressed in culturally unique ways. It is precisely this cultural uniqueness that should be preserved in the translation process. The aim is to make the translation of proverbs transparent, not in the sense of Venuti, but in the sense that it enables us to see through the target language the cultural aspects of the source language, hence making translation an educating tool and making the reading of translated literature a learning experience. Norrick rightly argues that proverbs are "not only linguistic items inventorised in some language, but also items of folklore traditional in some culture" (1985, p. 4). Only semantic translation can preserve this folkloric dimension and

Proverb Translation: Fluency or Hegemony?

67

appreciate its didactic role in intercultural communication, exchange and learning. Jin Di believes that "creative imagination is the hallmark of a work of art. In a way it is also what gets recognition first in a literary translation. That and the naturalness of style were perhaps the main reasons behind the success of Edward FitzGerald, Lin Shu and Ezra Pound....Those famous translators could...have produced still greater translations if they had respected their authors creative imagination more. The art of literary translation, more than anything, is the art of vivifying the author's creative imagination in a new language" (2003, pp. 88-89). Di is blaming these great translators for not being faithful enough to their authors and maintains that, had they been more respectful, their translations could have been even greater. This statement is even more relevant to proverb translation as, in this context, there seems to be little point in rendering proverbs communicatively or idiomatically using situational equivalents from the target language. In most cases, one’s own proverbs and cultural clichés will tend to get repeated. While it is true that the rendition will be smooth, clear and natural because the translation is domesticating something that is foreign and making it familiar to the recipient, it is precisely this cultural comfort that one should avoid, for the foreign proverb is being portrayed in ways that are too familiar to us. In addition to the fact that we are giving a false impression of the fluency of the foreign proverb by appropriating it, the proverb's important educational and instructional element will also be lost. Di's statement echoes Reiss' (2000) emphasis on preserving “the original author's external or inner forms in translation” (p. 35). When attempting to characterise the function of language in certain texts, Reiss comes to the conclusion that "language for practical texts is primarily a means of communication, of conveying information, while for literary prose and poetry it is a tool of artistic creativity, conveying aesthetic values" (2000, p. 18). This shows that semantic translation allows the TL reader to see the source culture through the TT, and this is indeed the cultural responsibility of the translator who should offer readers a transparent window to see through. Communicative translation, on the other hand, hides source cultural references and gives a false impression of fluency, thereby sacrificing source culture values for the sake of the smoothness and clarity of the TT. Venuti (1992) calls the semantic approach to translation a "resistant translation strategy" because it defies dominant perspectives and tries to depart from hegemonic trends that victimise source cultures and source languages. He maintains that "resistant strategies can help to preserve the

68

Chapter Four

linguistic and cultural difference of the foreign text by producing translations which are strange and estranging, which mark the limits of dominant values in the target-language culture and hinder those values from enacting an imperialistic domestication of a cultural other" (pp. 1213). The semantic approach advocated here finds its full support in the argument of the German theologian and philosopher Schleiermacher (1813) who proposed two options that a translator usually has. One is to bring the author to the reader, thereby leaving the reader in peace, or taking the reader to the author and leaving the author in peace. Schleiermacher advocates a foreignising method that sends the reader abroad. He champions foreignisation because, in his view, it is the proper translation strategy to bring readers closer to the linguistic and cultural world of the author. He suggests that a foreignising approach opposes the foregrounding of signifieds, the method always favoured by free translators. He emphasises the role of focusing on the denotative dimension of signifiers rather than their connotative dimension. This explains the focus in this paper on the semantic approach as it can help to question and debate the linguistic and cultural ‘plundering’ practiced by some masters of translation, as indicated by Karl Vossler (cf. Venuti, 1992, p. 196). Semantic translation can diminish ethnocentrism and combat tendencies towards intra-cultural connections because if we translate in ways natural to readers, then the element of learning from other cultures will disappear. This deliberate effacing of foreignness and of the educating value in proverb translation threatens the very essence of translation. This concern is clearly voiced by Milan Kundera who, according to Venuti (1998), "is rightly suspicious of domesticating translations that assimilate foreign literary texts too forcefully to dominant values at home, erasing the sense of foreignness that was likely to have invited translation in the first place" (p.5). Francis Newman, a 19th century classical scholar, saw in the semantic approach, a tool for signifying remoteness in both place and time and hence advocated a foreignising approach to literary translation. He "recommended a translation method that signified the many differences between the translation and the foreign text, their relative autonomy from one another, their composition in different languages for different cultures" (Venuti, 1995, p. 122). The theoretical points raised above can be further explained by examining the following two examples of proverbs translated from English into Arabic using the semantic approach. They are taken from (ODEP) and their semantic translations from Attia (2004).

Proverb Translation: Fluency or Hegemony?

69

(1a) Make hay while the sun shines. ΔόϟΎρ βϤθϟ΍ϭ ΐθόϟ΍ ϒϔΟ Transliteration: jaffif al ?ushb wa sh-shamsu taali?atun Back translation: Dry grass while the sun shines.

This is a fairly faithful translation because the same semantic elements are preserved and it is easy to look through the translated proverb to see the source culture's thought patterns which, in this particular example, are agricultural in orientation. Although the translator changed 'hay' to 'grass', the semantic field remains intact because it still relates to a farming culture. This kind of translation takes the Arab reader to the source context and enables them to experience a situation in which the sun comes out only occasionally. This, of course, is not the case in the Arab context where the sun is almost always shining; hence the cultural enrichment and educational component of the proverb. Another good example which shows how semantic translation retains source-cultural references and feeds them to the target culture, thus enriching the host culture, is given below: (2a) The proof of the pudding is in the eating. ϞϛϷ΍ ΪϨϋ ϻ· ϡΎότϟ΍ ΓΩϮΟ ϑήόΗ ϻ Transliteration: laa tu?rafu jawadatu at-ta?aami 'illaa ?inda al-'akli Back translation: the quality of the food is known only when it is eaten.

Even though the translation provided by Attia (2004) does not fully preserve the source culture because it replaced the word 'pudding' with a rather too general word 'ϡΎόρ' , i.e. 'food', it is still largely a semantic rendition because most of the source-cultural references are still intact. In other words, in both the source text proverb and the target text translation of that proverb, the image content is the same, i.e. food and tastes of food. This kind of translation leaves interpretation of the pragmatic meaning of the proverb to the reader, hence respecting both reader and proverb. The element of learning is also maintained in this kind of translation, although it could have been more significant if the word 'halwa', i.e. 'sweet' had been used instead of 'food' because this specification is necessary to reflect cultural preferences and modes of thinking.

70

Chapter Four

Communicative Translation, Cultural Apprehension and Cultural Hegemony Newmark (1981) defines communicative translation as an activity "where the translator attempts to produce the same effect on the TL readers as was produced by the original on the SL readers" (p. 22). Munday (2008) summarises Newmark's description of communicative translation as follows: "it is reader focused,..smoother, simpler, clearer, more direct, more conventional, appropriate for non-literary writing” (p. 45). Domesticating approaches to translation have long been used in Europe and the practice goes back to Cicero, Quintilian and others who emphasised that they translated from Greek into Latin using a sense-forsense, not word-for-word, approach, emphasising freedom rather than literalness. St. Jerome described his strategy in translating from Greek as follows: "Now I not only admit but freely announce that in translating from the Greek – except of course in the case of the Holy Scripture, where even the syntax contains a mystery–I render not word-for-word, but sensefor-sense” (cited in Munday, 2008, p. 20). Martin Luther's preferred approach was also a free translation which is captured in his typical quote: "You must ask the mother at home, the children in the street, the ordinary man in the market, and look at their mouths, how they speak, and translate that way; then they'll understand and see that you're speaking to them in German" (cited in Munday, 2008, p. 23). Venuti (1992) argues that "a translated text is judged successful–by most editors, publishers, reviewers, readers, by translators themselves– when it reads fluently, when it gives the appearance that it is not translated, that it is the original" (p. 4). He further maintains that "a fluent strategy effaces the linguistic and cultural difference of the foreign text…[it] performs a labour of acculturation which domesticates the foreign text, making it intelligible and even familiar to the target-language reader, providing him or her with the narcissistic experience of recognizing his or her culture in a cultural other, enacting an imperialism that extends the dominion of transparency with other ideological discourses over a different culture" (p. 5). This approach is especially appropriate for instrumental and commercial texts where the translator normally feels no cultural responsibility. However, in scholarly, artistic and literary works like proverbs, there is always an invested interest in documenting the source culture by foreignising the target language to show a remoteness in time and place.

Proverb Translation: Fluency or Hegemony?

71

It seems that the 20th century's information and technological revolution, global economic trends and a widespread concern with profit have all markedly influenced the field of translation. Thus, purely instrumental modes are encouraged in order to achieve the utilitarian goals of practical knowledge transfer rather than cultural awareness and exchange. Hence, there appears to be a broad preference for domesticated translations that aim to extract, from a literary text only the communicative purpose and little else. In this regard, the British translator, John Cohen, says that "twentieth-century translators, influenced by science-teaching and the growing importance attached to accuracy...have generally concentrated on prose-meaning and interpretation, and neglected the imitation of form and manner” (cited in Venuti, 1995, p. 6). In this kind of situation, the aim of translation is "to bring back a cultural other as the same, the recognizable, even the familiar; and this aim always risks a wholesale domestication of the foreign text, often in highly self-conscious projects, where translation serves as an appropriation of foreign cultures for domestic agendas, cultural, economic and political" (Venuti, 1995, pp. 18-19). The Victorian writer Matthew Arnold, in a critique of Newman's translation of Homer's Iliad, advocated a domesticating approach which would become dominant in Anglo-American circles until today. This view received support in the 1950s and 1960s from the American Bible translator Eugene Nida (1969), who argued for a dynamic equivalence that aimed to achieve the desired effect on target readers. I will now turn to a discussion of a communicative translation of each of the two above mentioned examples, provided by Ba'labki (2001). (1b) ΎϬϤϨΘϏΎϓ ϚΣΎϳέ ΖΒϫ ΍Ϋ· Transliteration: 'ithaa habbat riyaahuka fa ghtanimhaa Back translation: when your wind blows, take advantage.

The translation in (1b) fails to provide the cultural input and is merely a repetition of an equivalent target culture proverb supposedly conveying the same pragmatic meaning and neglecting everything else. The question is, who can guarantee that this rendition conveys the same natural intent or effect as the source text proverb in (1a) above? The cultural worlds of both proverbs differ widely. The translation in (1b) fails to attend to the two translation norms identified in the introduction, namely cultural preservation and cultural enrichment because, it neither preserves the source-cultural references nor enriches the target culture with new modes of thinking and new cultural knowledge. While the semantic translation retains the farming culture references, the communicative translation

72

Chapter Four

sacrifices these for the sake of promoting target culture images and thoughts that are, at least in the translation context, well known, commonplace and ordinary. Although a domesticated translation looks natural and feels smooth, it can, in fact, be seen as an act of ‘intracultural inbreeding’ where the proverbs of the target culture are simply repeated, hence undermining the essence of translation as a cultural bridge that takes one across the river, and not merely as a communication bridge that enables an understanding of the Other. The same argument about source culture repression and target culture deprivation can be reiterated here with the following rendition which is a communicative adaptation of proverb (2a) above. (2b) ϥΎϬϳ ϭ΃ ˯ήϤϟ΍ ϡήϜϳ ϥΎΤΘϣϻ΍ ΪϨϋ Transliteration: ?ind al 'imtihaan yukramu l mar'u 'aw yuhaan Back translation: When tested, a man is either honored or shunned

The semantic worlds of the proverbs do not match. It is a world of 'pudding' and 'tasting' in the ST, but a world of 'tests', 'honouring' and 'humiliating' in the TT. The fact that the domestication strategy adopted here prevents target readers knowing what cognitive mechanisms source culture people associate with trying something and passing a judgment on the result cannot be disputed. If we translate the example in this idiomatic way, the reader will have access only to their mother tongue form and the proverb's meaning and intention, but will learn nothing new, although reading, understanding and appreciation will be easy. Appreciation, however, results from the fact that the reader is reading something made to look familiar to them, something whose estranging elements have been suppressed. But if the proverb is translated semantically, readers can learn how the source culture people express their meaning cognitively and culturally. TT readers will, even if they are monolingual, have an opportunity to make comparisons between what they have in their own culture and what other people have in theirs. This, it appears, is essentially the function of translation.

Conclusion This paper has argued for an approach to translation, which is called 'semantic'. This semantic approach aims to preserve more of the source culture and enrich the target culture. The semiotic definition proposed enables proverbs and their translations to be seen as systems comprising three levels of signification. The paper's focus has been on the semantic level of signification, of both, proverbs and their translation, because this

Proverb Translation: Fluency or Hegemony?

73

is believed to display the potential aspect of meaning, leaving room for interpreters to employ their own mechanisms to suit the context. In other words, this semantic level of signification is transcendental in the sense that it is context-free and it would thus be possible to encode source culture meanings and feed them into the target culture.

Bibliography Abu Ssaydeh, A. 2004. Translation of English Idioms into Arabic. Babel, 50 (2), pp. 114-131. As-Safi. A. and I. ash-Sharifi. 1997. Naturalness in Literary Translation. Babel 43: (1), pp. 60 – 75. Attia, M. 2004. Dictionary of Common English Proverbs: Translated and Explained. [ebook retrieved 1 February 2010. from http://www.attiaspace.com/]. Ba'labki, M. 2001. Al-Mawrid English – Arabic Dictionary, Beirut: Daar al Ilm lil malaayeen. Berman, A. 2004. Translation and the trials of the foreign. In Venuti, ed. 2004. Eco, U. 1984. Semiotics and the Philosophy of Language. London: Macmillan. El-Yasin, M. and A. Al-Shahabat. 2005. 'Translating Proverbs'. Babel, 51 (2), pp. 161-173. Frege, G. 1966. 'On Sense and Reference', in Geach, P. and M. Black eds. and trans. Translations from the Philosophical Writings of Gottlob Frege. Oxford: Blackwell. House, J. 2001. Translation Quality Assessment: Linguistic Description versus Social Evaluation. Meta, 46 (2), pp. 243 – 257. Jin, Di. 2003. Literary Translation: Quest for Artistic Integrity. Manchester: St. Jerome. Munday, J. 2008. 2nd ed. Translation Studies: Theories and Application. London: Routledge. Newmark, P. 1981. Approaches to Translation. Oxford and New York: Pergamon. Nida, Eugene. 1969. Science of Translation. Language vol. 45 no. (3). Norrick, N. 1985. How Proverbs Mean: Semantic Studies in English Proverbs. Berlin: Mouton. The Oxford Dictionary of English Proverbs. 1970. Oxford: Oxford University Press

74

Chapter Four

Peirce, C. S. 1960. Collected Papers of Charles Sanders Peirce, vols. 1 – 8, eds. Hartshorne, C. and P. Weiss. Cambridge: Harvard University Press. Reiss, K. 2000. Translation Criticism – The Potentials & Limitations. trans. F. Erroll. Rhodes. Manchester: St. Jerome. Turner, G. 1992. British Cultural Studies: An Introduction. New York: Routledge. Venuti, L. 1992. ed. Rethinking Translation: Discourse, Subjectivity, Ideology. London: Routledge. —. 1995. Translator's Invisibility: A History of Translatson. London: Routledge, —. 1998. The Scandals of Translation: Towards an Ethics of Difference. London: Routledge. —. 2004. ed. The Translation Reader. London: Routledge. Wilson, J. 1970. Introduction to the Oxford Dictionary of English Proverbs. Oxford: Oxford University Press.

CHAPTER FIVE DIALOGUE SYSTEMS: STRETCHING THE BOUNDARIES OF PRAGMATICS AND DISCOURSE ANALYSIS RADHIKA MAMIDI

Abstract The aim of every computational linguist is to make the computer interact with human beings in a natural way and display intelligence by understanding the complex cognitive structure of language. In this paper, features of human conversation and how these help in understanding the structure of conversation are discussed. Using samples of language use, the paper demonstrates how we naturally use certain pragmatic and discourse techniques such as indirect speech acts, presuppositions, implicatures and elliptical sentences in our conversations. We may also easily recognize the speaker’s intention, draw inferences, resolve references of the pronouns etc. To do this, we make use of the available contextual knowledge and world knowledge. When computers are expected to talk with human beings, they need to be able to handle such pragmatic and discourse techniques. This paper attempts to stretch the boundaries of pragmatics and discourse analysis studies related to human-human dialogues to encompass human-computer dialogues. The focus of this paper is to discuss various pragmatic and discourse issues that need to be handled in the language understanding module of a dialogue system.

Keywords: pragmatics, discourse analysis, conversation, machines, dialogue systems, computational models

Chapter Five

76

Introduction A dialogue is considered as a two-way dynamic communication between speaker and hearer, where the roles of speaker and hearer keep interchanging. A dialogue is coherent when the participants share the textual, contextual and world knowledge. Studies have shown that conversations have a structure and the mechanism of turn-taking is systematic (Sacks et al, 1974; Schegloff, 1972; Levinson, 1983; Green, 1987). The simple 2-line dialogues given below give sufficient insight into issues related to pragmatics and discourse. (i)

A: Hi! How was your holiday? B: With in-laws. What do you think it was like? (ii) A: Hi! How was the movie? B: It was good. (iii) A: How come you are at home? B: Today is Friday.

These dialogues show that knowing just the meaning of words or having knowledge of syntactic structures is not sufficient enough to understand each other. For instance, to understand B, A should know what It in (i) B and (ii) B refers to. In the same way, A should know the implication in (iii) B which is that schools/offices are closed on Fridays (in the Middle East) and so B is at home. Knowing the intention of the speaker and making correct inferences are very important as well, as seen in understanding the meaning of the phrase “It is cold” below. (iv) A: It is cold. B: Shall I turn off the air conditioner? (v) A: It is cold. B: Shall I heat it up in the microwave oven?

The phrase “It is cold” does not carry the same pragmatic force in (iv) and (v). The intention of speaker A was different in each context and B inferred the intention correctly. A difference also lies at the discourse level—‘It’ in (iv) is an empty pronoun, whereas ‘It’ in (iv) refers to some exophoric noun phrase (which is not mentioned in the text)—a dish, perhaps. This difference can be discerned by B’s response in each case. This level of language analysis is a challenge for machines. Some more issues like these will be discussed in subsequent sections. The issues related to morpho-syntactic levels will not be taken up in this paper as they

Dialogue Systems

77

are more easily handled, compared to pragmatic and discourse levels of analysis and generation in computational linguistics. All the samples of typical human-human dialogues and other examples used in this paper were collected from different sources or constructed for the author’s ongoing work in this area, unless otherwise cited.

Dialogue Systems Dialogue Systems is the most important application area of Computational Pragmatics which is a sub-field of Computational Linguistics. Computational Pragmatics draws on disciplines such as Pragmatics, Artificial Intelligence and Cognitive Science to study the research, development, implementation, and evaluation of dialogue systems or conversational agents. Studies in pragmatics related to speechacts, indirect speech-acts, conversational implicatures etc. by exponents like Austin (1962), Searle (1969) and Grice (1975) form the basis of Computational Pragmatics. Much progress has been made in the field of Dialogue Systems since Eliza, the first chatbot, was developed by Joseph Weizenbaum in the late 1960s (Weizenbaum, 1966). Eliza was based on Rogerian Psychotherapy and used pattern matching to carry on a limited form of conversation. It gave a feeling that it was ‘human’ as its answers showed its ‘intelligence’ in processing the input and generating suitable outputs. Today, there are much more sophisticated conversational agents available online serving a wide range of purposes. For example, many companies have chatbots serving as assistants and consultants in education, business, travel, hospitality, healthcare, and entertainment sectors. In recent times, one of the most successful applications of dialogue systems has been with in-car navigation. The machine helps the user not only by giving directions but also in making calls or choosing music. The following dialogue between User (U) and System (S) from CSLI Dialogue System taken from Pon-Barry et al (2004) for in-car devices gives a good idea of the extent of the usefulness of spoken dialogue systems. 1U: 2S: 3U: 4S: 5U: 6S: 7U: 8S:

What music do you have? I have rock, soul … What rock songs? Just a second: you have to turn left at the next intersection. OK. I have songs by U2, Cake, Green Day, Jet, … Play something by them. OK. Playing xxx.

Chapter Five

78 9U: 10S: 11U: 12S: 13U: 14S: 15U: 16S:

Call John Do you mean John Smith or John Barry? John Smith Calling… I need somewhere for lunch. There is a highly rated vegetarian restaurant near here. What other cuisines? There are Italian, Japanese, Malaysian, Nepali, and Thai restaurants close by.

This particular spoken dialogue system has access to the car’s internal databases such as the phone directory and the music playlist, and also has access to several web-based applications such as Google maps. Research is continuing in several places across the world in building dialogue systems in different languages. Depending on the use and mode, the different systems available can be classified into the following types: x x

x x x x x

Interactive Voice Systems (e.g., railway enquiry system asking the user to press certain numbers for accomplishing the task) Question-Answer Systems (e.g., START, a web-based question and answer system answering user's questions on movies, places, people etc.), Natural Language Interfaces (e.g., tutoring systems and trip planning systems) Personal Assistants that understand user needs and tasks Intelligent Tutoring that involves student engagement, Embodied Conversational Agents that engage via realistic and affective physical and facial gestures, and Intelligent Systems for differing environments, such as the home or car, that understand user situation and activity.

An alternative way of classifying dialogue systems is based on initiation techniques: x x

single initiative systems where the machine guides users through a series of scripted prompts as in telephone enquiry systems, or mixed initiative systems where both the users and machine can initiate the conversation in a cooperative and flexible way as in flight booking or car navigation.

A single initiative system limits the interaction, but the language processing is much better than in a mixed initiative system as the users’ responses are predictable. However, researchers are more interested in developing mixed-initiative systems as they are closer to the structure of

Dialogue Systems

79

human conversation. Doran et al (2001) observed that conversation was more balanced in human-human situations in terms of the amount of speech, types of dialogue acts and with respect to initiative, while the system dominated in number of words and dialogue acts and in initiative in human-computer conversations. To understand dialogue systems better, I will draw on the architecture of a simple spoken dialogue system developed for the domain of railway reservations (Mamidi and Khan, 2005). Language Understanding (LU)

user utterance

Automatic Speech Recognition (ASR)

Language Generator (LG)

Dialogue Manager

Text to Speech (TTS)

spoken output

Domain Manager

Figure 11.1: Architecture of a simple dialogue system

As can be seen in figure 11.1, the system has six modules, with the dialogue manager being the hub of the system. The dialogue manager plays an important role in interacting with the other modules. When the user utters a sentence, it is converted into text by the automated speech recognition (ASR) system. The text is given to the language understanding (LU) module where all the language specific tasks are completed. The output of this module is rich with linguistic information which helps the dialogue manager in proceeding to take the next step. For instance, it helps the dialogue manager to figure out the intention of the speaker and then draw appropriate inferences. Similarly, other tasks like ellipsis handling and anaphora resolution are dealt with in this module. The domain manager gives the dialogue manager specific information based on the user’s needs. Finally, the dialogue manager interacts with the language generator module to generate natural language sentences which the text to speech (TTS) module converts into spoken sentences. The issues related to automated speech recognition (ASR), text to speech (TTS) or natural language generation modules will not be discussed here as the focus of the

Chapter Five

80

present paper is on the language understanding (LU) module which will be discussed in the following sections.

Language Understanding Module The task of the language understanding module is to process the output given by the ASR module. After the language processing is completed, the output representing the phonological, morphological, syntactic, semantic, pragmatic and discourse information is given to the dialogue manager for further operations. This module is the most challenging one as human language is complex and ambiguous. To understand language, different techniques, apart from linguistic knowledge and general knowledge, are needed. Examples like the following (taken from different sources including newspapers and everyday conversation) demonstrate the knowledge required by machines to process each one of these inputs when the context is missing. (i) (ii) (iii) (iv) (v)

A nameless man/ An aimless man (speech to text problem) She showed me the mouse. (mouse = rodent or computer part) The leopard was spotted. (spotted = adjective or verb) It is not my cup of tea. (cup of tea = literal or figurative) If the boys don’t like the new toys, throw them away. (them refers to the boys or new toys)

The examples above (i-v) reflect ambiguity at phonological, lexical, syntactic and discourse levels that can be resolved using contextual knowledge. The examples below (vi-vii) have the context included to show that we process language easily only because we have internalised all the requisite knowledge. (vi) When shot at, the dove dove into the bushes. (Correct pronunciation of ‘dove’ according to the part of speech) (vii) Upon seeing the tear in the painting I shed a tear. (Correct pronunciation of ‘tear’ according to the meaning)

It is interesting to see how humans process language so easily. Greene (1986, p. 96) presents a cooperative model of language processing (figure 11.2) showing the interactions between the various types of knowledge. As shown in the figure below, there is an interaction between two sources of knowledge: the input consisting of words and phrases and the information stored in one’s memory. Then there are processes (lexical,

Dialogue Systems

81

syntactic, semantic and discourse) that interact with each other. The information is pooled from these different resources to generate an appropriate output.

Memory General Knowledge Lexicon

inputs

Lexical Processing

Syntactic Rules

Syntactic Processing

Semantic Rules

Discourse Rules

Semantic Processing

Discourse Processing

outputs

Figure 11.2: Interactive model of language processing

We can understand the model if we take an example phrase, such as The internet links. Depending on the words that follow, the meaning of the phrase changes as seen in The internet links us to the world or The internet links are given below. This shows that for language processing, interaction with all the available knowledge viz. lexical, syntactic, semantic and discourse is required simultaneously. This model is more cognitively efficient than a linear model which prefers language processing in a sequential manner—first lexical, then syntactic and so on. For example, a linear model will analyse the first word The first, then The internet and then The internet links. If links is marked as a noun and the following word is us, then it is a wrong analysis. In the interactive model, there is no scope for such errors as it can look-ahead before analysing the current word. The interactive model is the most suitable one for the language understanding module in a dialogue system.

Some Pragmatics and Discourse Issues As discussed earlier, we need several kinds of knowledge to understand each other—linguistic knowledge, contextual knowledge and world knowledge. When we lack full information (contextual knowledge)

Chapter Five

82

misunderstandings arise. But humans have a way of resolving this through more information, as can be seen from the example below: (i)

Mother: Reem called you. Daughter: Oh, it must be about tomorrow’s test. Mother: She mentioned something about the cookery class. Daughter: Oh! It must be Reem Al-Saby! Let me call her.

From this conversation, it is clear that the daughter knows two Reems and the clue was found in the phrase ‘cookery class’ which resolved the correct referent of the referring expression ‘Reem’. Sometimes, if the hearer lacks the contextual knowledge, the speaker’s role requires providing this to the hearer. For example, in the following conversation, B provides C with the necessary information to understand the ongoing conversation between A and B. (ii)

A: Oh, lovely curtains. Ikea? B: Yes. They cost a fortune. C: What is Ikea? B: It is a store that sells household things... A: And much more. You should visit it once.

The following are some of the implicit notions in pragmatics, and implicit and explicit notions in discourse which need to be handled in a dialogue system.

Implicit Notions in Pragmatics A machine will find it difficult to detect the following notions as they are not represented explicitly in text. Presupposition Presuppositions are inferences drawn from an utterance. They are assumed by the hearer and not asserted directly by the speaker. (i) (ii)

I want to buy a ticket for my mother as well. (It is presupposed that I have already bought one ticket) Cancel my mother’s ticket to London. (It is presupposed that my mother bought a ticket to London).

Intention The intention of the speaker has to be understood properly for successful communication. Failure in judging the intention of the speaker

Dialogue Systems

83

may lead to misunderstanding or miscommunication. The same utterance in different contexts may reflect the intention of the speaker differently. For example: (i) I bought two tickets to Nice. They are expensive. (Assertion) (ii) I cannot buy these tickets. They are expensive. (Reason) (iii) Be careful when you pack these. They are expensive. (Warning)

Implicature When the speaker violates or flouts any of the maxims formulated by Grice (1975), the hearer has to draw implicatures. (i) (ii)

A: Have you finished your packing? B: I have just started it. (implicature = I have not finished it) A: You should finish it soon or you may get late. A: Are your parents coming to see you off at the airport? B: Mom is coming. (implicature = only mother is coming) A: I guess your Dad is busy.

In both the situations, A could correctly detect what B implied and so responded appropriately.

Implicit notions in Discourse Along with the above mentioned pragmatic issues, one needs to look at issues related to coherence in monologue or dialogue. In other words, for a text to be coherent, the sentences should be related. The link between the sentences can be seen from anaphors, discourse relations and discourse connectives. Discourse Relations Sometimes, discourse relations such as elaboration, reason, contrast, condition, result, sequence, explanation etc. between sentences are difficult to identify when the complete context is missing. For example, in each of the following, notice the relationship which the second sentence has with the first sentence: (i) (ii)

Jane was angry. The flight got delayed. Jane was angry. John asked her to relax.

Though the structures are similar, in the first case, the second sentence is an explanation and in the second case it is a sequence of actions. This information is not seen explicitly. Mann and Thomson (1987) propose

Chapter Five

84

Rhetorical Structure Theory (RST) to capture the relations between sentences in a discourse. Though a dialogue has a structure different from a monologue, information to make links and sense across sentences is still needed. Discourse Connectives Connectives, sometimes referred as cue phrases, may give a different meaning to discourse. For instance, connectives such as but, and, so and because inserted between the two sentences below indicate different discourse relations. (i) (ii)

John likes to travel in summer. Mary loves to travel in winter. John pays by cash. His brother uses Visa credit card.

The second sentence might be a sequence, reason, contrast or explanation depending on which cue was used viz. but, and, so and because. When the cues are implicit, then identifying them correctly becomes a challenge. A solution for this is to program the machine to learn from a corpus annotated with the discourse connectives using statistical techniques.

Explicit notions in Discourse Anaphora Resolution Anaphora resolution is the identification of the correct antecedent of a given pronoun. Jurafsky and Martin (2000) present in detail the different types of anaphors and the different algorithms given by Lappin and Leass (1994), Hobbs (1978) and Grosz et al (1995) to identify the noun phrases the pronouns refer to. Unless the pronoun is resolved, it is difficult to carry out a conversation. The following are examples where B in each case has resolved the identification of the correct pronoun. (i) (ii)

A: Jane persuaded Mary to go to her house. (her house refers to Jane’s house) B: Did Mary go? A: Jane promised Mary to go to her house. (her house refers to Mary’s house) B: Did Jane go?

In these examples, the resolution depends on the verb. The following examples show the role of expectation in anaphora resolution. The notion of expectation requires world knowledge associated with the verbs.

Dialogue Systems (i) (ii)

85

A: John hit Bill. He was scolded. (He = John or Bill) B: John always bullies Bill. A: John hit Bill. He started crying. (He = John or Bill) B: Bill is such a sissy.

Other factors like number, person, gender and case agreement, are taken into account while writing instructions to programmers for anaphora resolution. Ellipsis In a text, the structure of the previous sentence helps in reconstructing the elliptical sentences. (i) (ii)

Mary booked her tickets. So did Susan. Mary bought a souvenir. So did Susan.

In (i) the newly constructed, elliptical sentence is “Susan booked her tickets” and in (ii) it is “Susan bought a souvenir”. A sentence generator takes the output of a rule based parser ran on the first sentence and then substitutes the noun phrase of first sentence with the noun phrase of the second sentence to construct the elliptical sentence.

Issues unresolved computationally Depending on the design requirements of the system, researchers have written new algorithms to solve differing linguistic issues or have adapted the algorithms written for specific applications. There are, however, some issues that remain unresolved, such as the techniques we use for sarcasm and humour in our conversation. There are also other issues related to kinesics and paralanguage which are important in communication, but are not handled computationally. For example, humans use many paralinguistic cues like hums, grunts, pauses, false starts, hedges, disfluencies, hesitations, temporal and prosodic features in their conversations (cf. Chung, und.; Carlson et al, und.; Pon-Barry, 2004.; Weng et al, und.; Doran et al, 2001.). Some of these cues may have a linguistic meaning as well. A computational model for these features is still not available.

Chapter Five

86

Computational Models Currently, the available dialogue systems are for specific, limited domains. Building a general purpose dialogue system is far more challenging, as all the pragmatic and discourse issues, discussed earlier, need to be solved. The following figure (11.3) from Allen et al (2001) summarises some models that use different techniques such as finite-state (where the structure is fixed and dialogue acts are represented as adjacency pairs, e.g., greeting-greeting, question-answer), frame-based (where the machine interacts in such a way that a template gets filled up) and plan-based techniques (where a plan may be decomposed into sequences of actions and sub-goals) to accomplish tasks such as booking tickets, enquiring about a product, designing house or helping in saving lives through the efficient use of information technology during a disaster. Figure 11.3 also shows the range of systems currently available. All of the systems shown in Figure 11.3 can be classified as practical, goal-oriented dialogue systems with the structure of the dialogue being either task-oriented or information seeking. It is a fact that goal-oriented dialogue is simpler to implement than the general-purpose dialogue (Allen et al, 1998). The ultimate challenge is to devise a dialogue system that will not be limited by either domain or task. The dialogue systems built on the techniques mentioned above work well for limited domains such as travel, sports, tourism, health, etc. For any system to be judged intelligent, it should have the components given by Allen (1995). These are: x x x

x x x x

Perception: the agent must be able to perceive the world around it Beliefs: the agent must have a representation of the present state of the world Desire/wants: the agent should have positive or negative responses to various states of the world, creating a way to compare the desirability of states Planning/reasoning: the agent must be able to reason about ways to attain other states Commitment: the agent must be able to decide to act to get to a different state Intentions: the agent must be able to maintain the course of action decided on Acting: the agent must be able to act and thus change its state

Dialogue Systems

TECHNIQUE USED

EXAMPLE TASK

TASK COMPLEXITY

87 DIALOGUE PHENOMENA HANDLED

Least complex Finite-state script

Long-distance dialling

User answers questions

Frame-based

Getting train timetable information

Set of contexts

Travel booking agent

User asks questions, simple clarifications by system Shifts between predetermined topics

Plan-based models

Kitchen design consultant

Agent-based models

Disaster relief management

Most complex

Dynamically generated topic structures, collaborative negotiation subdialogues Different modalities (e.g. planned world and actual world)

Figure 11.3: Dialogue modelling frameworks

Most of these features are seen in human conversation also. For successful conversation, even of a human-human type, recognizing the belief, desire and intention, all pertaining to the cognitive states, is important. A model along these lines is called the BDI (belief, desire and intention) model. It is an extremely powerful approach to dialogue act comprehension and speech act interpretation (Jurafsky and Martin, 2000). It uses rich knowledge structures and powerful planning techniques. It addresses even subtle indirect uses of dialogue acts. It incorporates knowledge about the speaker’s and hearer’s intentions, actions, knowledge, and belief, that is essential for any complete model of dialogue. However, this model has its limitations. It requires that each utterance has a single literal meaning, which is considered by plan inference rules to produce a final non-literal interpretation. An alternative model is cue based model (Jurafsky and Martin, 2000). This model is based on the assumption that the listener uses different cues in the input to help decide how to build an interpretation. The surface input to the interpretive algorithm provides clues to structure-building, rather than providing a literal meaning which must be modified by purely inferential processes. The different sources of knowledge (cues) such as lexical, collocational, syntactic, prosodic, or conversational-structure cues are used in the cue-based model for detecting a speech act.

Chapter Five

88

Both these models generally use machine learning algorithms. For this purpose, a corpus of dialogues annotated with dialogue acts is required. Dialogue acts are speech acts marked with additional conversational functions and pragmatic force. The following section discusses dialogue acts and its role in dialogue systems.

Representing Speech Acts as Dialogue Acts An important resource to handle issues higher than the syntactic level is to have a corpus annotated for dialogue acts. Currently, much effort is going on in the annotation of corpora of dialogues. The machines are trained on the annotated corpora before they are developed into automated systems. A currently available tagging scheme is the DAMSL, Dialogue Act Markup in Several Layers, scheme. DAMSL is a dialogue act tagging scheme developed especially for task-oriented dialogue (Allen and Core, 1997). By taking into account the content of the utterance and the intention of the speaker, the tags are classified into four main categories: (i)

Communicative Status: it records whether the utterance is intelligible and whether it was successfully completed. (ii) Information Level: a characterization of the semantic content of the utterance. (iii) The Forward Looking Function: how the current utterance constrains the future beliefs and actions of the participants, and affects the discourse. (iv) The Backward Looking Function: how the current utterance relates to the previous discourse. (Allen and Core, 1997)

The following conversation marked with the dialogue acts [in square brackets] is an example of an annotated dialogue. A: B: A: B: A: B: A: B: A:

I think we should meet soon. [Suggestion] Yeah. [Agreement] How about (pause) tomorrow? [Suggestion] Hmm. Tomorrow I am busy. [Negation] Whole day? [Negotiation begins] Hmm. I have a class from 10 to 12 and then 2 to 4 a long meeting. [Elaboration] How about at 12? [Insistence/Suggestion] (Snigger) Lunchtime? [Dislike the option] Yeah. (Smile) [Affirmation]

Dialogue Systems B: A: B:

89

Okay then. [Agreement] Don’t worry. We will be back on time. I know this restaurant across the road... [Assurance/Suggestion] Sounds good.

A machine trained on such annotated corpora should be able to conduct a successful conversation with humans. A computational linguist aims for the following type of conversation between a User (U) and a System (S). This conversation is based on a Wizard of Oz type of experiment for a prototypical flight reservation domain. The pragmatic and conversational techniques used are given in comments. U1: Please book a ticket to London. //Violation of maxim of Quantity (less information is provided)// S1: From Riyadh? //Information seeking// U2: Yes. And one to Paris during Hajj vacation. //Ellipsis (one means one ticket)// S2: You need to book one ticket to London and one ticket to Paris during Hajj vacation. //Seeking confirmation after ellipsis handling and context handling from previous turns// U3: Yes. //Confirmation// S3: When do you plan to travel? //Information seeking// U4: Well, my first day of conference falls on 10 October. //Violation of maxim of clarity// S4: Do you want to travel on 9 October? //Prediction// U5: What is the flight’s arrival time in London? //Query// S5: On 9th October, British Airways has a flight reaching London at 5pm. Do you want me to check other airlines? //Options offered// U6: No. This is perfect. //Confirmation// S6: Okay. One ticket is booked on 9 October from Riyadh to London on British Airways. //Task completion// U7: Good. //Confirmation// S7: You want to make more bookings? //New task//

As can be seen, techniques related to conversation analysis (turntaking, adjacency pairs), discourse analysis (ellipsis) and pragmatics (speech acts) as well as the integration of linguistic cues into computational models collaborate in making a successful dialogue.

Conclusion Dialogue is one of the most important functions of human language. In the field of dialogue systems, dialogues play an important role by allowing us to explore the complex nature of human language and at the same time

90

Chapter Five

test novel challenges in language technologies. At present, there are dialogue systems available for limited domains only, but the desired goal has always been to build robust dialogue systems for mixed-initiative, multi-domains and multi-party interactions. The challenges in building general purpose systems are mainly in respect of pragmatic and discourse levels. Some of the issues that are to be handled are recognizing the speaker’s intention, drawing inferences, handling presuppositions and ellipses, resolving anaphora, identifying discourse relations and so on. By extending the findings in pragmatics and discourse studies, researchers are trying to resolve these issues and develop good computational models to build robust dialogue systems. Currently, computational models such as cue-based and BDI (belief, desire, and intention) require annotated dialogue corpora which help in the automatic identification of dialogue acts. This, in turn, will help machines understand the complexities of language at pragmatic and discourse levels to carry out successful conversations.

Bibliography Allen, James. 1995, Natural Language Understanding. CA: Benjamin/ Cummings Publishing Company, Inc. Allen, James and Core, M. 1997. Draft of DAMSL: Dialog act markup in several layers. Retrieved from http://www.cs.rochester.edu/research/ speech/damsl/RevisedManual/RevisedManual.html on 1 January 2010. Allen, James, D. Byron, M. Dzikovska, G. Ferguson, L. Galescu & A. Stent. 1998. An architecture for a generic dialogue shell. Natural Language Engineering. Cambridge: Cambridge University Press. Allen, F. J., Byron, D.K., Dzikovska M, Ferguson G, Galescu G., Stent A. 2001. Towards Conversational Human-Computer Interaction. AI Magazine. Austin, J.L. 1962. How to do things with words. Oxford: Clarendon Press. Carlson, R., J. Edlund, D. House, M. Heldner, A. Hjalmarsson, G. Skantze. Undated. Towards human-like behaviour in spoken dialog systems. Retrieved from http://www.speech.kth.se/prod/publications/ files/ 1789.pdf on 1 January, 2010. Chung, Grace. Undated. Developing a Flexible Spoken Dialog System Using Simulation. Retrieved from http://acl.ldc.upenn.edu/P/P04/P041009.pdf on 1 January, 2010. Doran, Christine, John Aberdeen, Laurie Damianos & Lynette Hirschman. 2001. Comparing Several Aspects of Human-Computer and HumanHuman Dialogues. Proceedings of the Second SIGdial Workshop on

Dialogue Systems

91

Discourse and Dialogue. Retrieved from: http://www.aclweb.org/ anthology/W/W01/W01-1607.pdf on 1 January 2010. Green, Georgia M. 1987. Pragmatics and Natural Language Understanding. New Jersey: Lawrence Erlbaum. Greene, Judith. 1986. Language Understanding: A Cognitive Approach. Buckingham and Philadelphia: Open University Press. Grice, H.P. 1975. Logic and conversation. In Cole & Morgan (eds.) Syntax and Semantics 3: Speech Acts. New York: Academic Press. Grosz, B. J., A.K. Joshi, and S. Weinstein. 1995. Centering: A framework for modelling the local coherence of discourse. Computational Linguistics, 21 (2), pp. 203-225. Hobbs, J.R. 1978. Resolving pronoun references. Lingua, 44, pp. 311-338. Jurafsky, D and J. Martin J. 2000. Language and Speech Processing: An Introduction to Natural Language Processing, Computational Linguistics and Speech Recognition. New Jersey: Prentice Hall. Available Online: http://www.cs.colorado.edu/~martin/slp.html Lappin, S and H. Leass. 1994. An algorithm for pronominal anaphora resolution. Computational Linguistics, 20(4), pp. 535-561. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Mamidi, R and Monis Khan. 2005. Linguistic Issues in Building Dialogue Systems. Hyderabad: LSI Platinum Jubilee Conference, 6-8 December, 2005, Mann, W.C. and S. A. Thomson. 1987. Rhetorical Structure Theory: A Theory of Text Organisation. Tech. Rep. RS-87-190. Information Sciences Institute. Available at http://www.sfu.ca/rst/ index.html Pon-Barry Heather, B. Clark, K. Schultz, E. Owen Bratt, and S. Peters. 2004. Advantages of Spoken Language Interaction in Dialogue-based Intelligent Tutoring Systems. Retrieved from http://www-csli. stanford.edu /semlab-hold/muri/papers/ITS_2004.pdf on 1 January, 2010. Sacks, H., E. A. Schegloff and G. Jefferson. 1974. A simplest systematics for the organization of turn-taking in conversation. In J. Schenkein (ed.) Studies in the Organization of Conversational Interaction. New York: Academic Press. Schegloff, E.A. 1972. Sequencing in conversational openings. In J. Gumperz and D. Hymes (eds.) Directions in Sociolinguistics. New York: Holt, Rinehart & Winston. Searle, J. R. 1969. Speech Acts. Cambridge: Cambridge University Press.

92

Chapter Five

Weizenbaum, J. 1966. ELIZA—A computer program for the study of natural language communication between man and machine. Communication of the ACM, 9(1), pp. 36-45. Weng, Fuliang, L. Cavedon, B. Raghunathan, D. Mirkovic, H. Cheng, H. Schmidt, H. Bratt, R. Mishra, S. Peters, L. Zhao, S. Upson, L. Shriberg, C. Bergmann. Undated. A Conversational Dialogue System for Cognitively Overloaded Users. Retrieved from http://godel.stanford .edu/ twiki/ pub/ Public/SemlabPublications/ weng-et-al04interspeechsystem.pdf on 1 March 2010.

CHAPTER SIX THE ROLE OF FORENSIC LINGUISTICS IN CRIME INVESTIGATION ANNA DANIELEWICZ-BETZ

Abstract This paper considers the extent to which forensic linguistics can be considered a science, and outlines some ways in which it is useful in legal proceedings, including voice identification, the interpretation of policesuspect interaction, verification of police reports (including the illegal practice of ‘verballing’) and cross-cultural insights into speech patterns in a courtroom context. The paper provides a closer examination of one particular area, that of authorship attribution, particularly in SMS messages, and concludes by raising some ongoing controversies in forensic linguistics and by discussing future prospects.

Keywords: forensic linguistics, authorship detection, authorship attribution, voice identification, forensic text types

Forensic Linguistics: An Introduction Forensic linguistics, as an emerging sub-discipline of forensic science, is an interdisciplinary field of applied/descriptive linguistics which comprises the study, analysis and measurement of language in the context of crime, judicial procedures or disputes in law. The interface between language, crime and the law can be detected, for instance, in the analysis of courtroom discourse, courtroom interpreting and translating, the readability/comprehensibility of legal documents, the comprehensibility of the police caution issued to suspects, and authorship attribution. Although it is, at present, far from being as accurate as DNA testing, forensic linguistics uses the expertise of descriptive and applied linguists

94

Chapter Six

in the unravelling of legal puzzles, so to say. Informed use of forensic linguistics requires familiarity with the broader application of linguistics as a social science, including phonetics and phonology, morphology, syntax, and semantics, discourse analysis, pragmatics, psycholinguistics, neurolinguistics, sociolinguistics, dialectology, computational linguistics, and corpus linguistics. The forensic linguist applies linguistic knowledge and techniques to the language implicated in legal cases or proceedings and private disputes between parties which may result in legal action. In this paper, I wish to first consider the extent to which forensic linguistics can be considered a science. I will then provide an overview of some of the areas in which forensic linguistics has a significant role to play—including voice identification, interpretation of police-suspect interaction, verification of police reports and cross-cultural insights into speech patterns—before turning to a closer examination of one particular area, that of authorship attribution. I will conclude by raising some ongoing controversies in forensic linguistics and discuss future prospects.

Is This A Science? The primary difference between forensic and non-forensic methods in linguistics is the scientific approach. In forensic linguistics, the scientific method requires hypothesis testing and a litigation-independent testing of the method for its accuracy. These tests are performed with robust controls regarding data quantity, data sources, and analytical objectivity. Restrictions in applying linguistic expertise in the context of law are due to varying degrees of acceptability in the courtroom, varying degrees of reliability related to shortcomings such as the brevity of documents, small data samples, general characteristics of language (for example, generic language features of suspects), and the intrinsic nature of language as something in constant change. The quality of evidence from this emerging field also depends considerably on the experience and knowledge of individual linguists involved in a given case. Courts in many countries admit forensic evidence but have differing criteria. In the United States, for example, the so-called Daubert standard rule of evidence regarding the admissibility of expert witnesses’ testimony in federal legal proceedings states that evidence based on innovative or unusual scientific knowledge may only be admitted after it has been established that it is reliable and scientifically valid. The Daubert test is based on peer review, error rates, testing, and acceptability in the relevant scientific community.

The Role of Forensic Linguistics in Crime Investigation

95

Is there a linguistic equivalent of an individual fingerprint—a ‘linguistic fingerprint’? This is indeed an attractive notion, which would certainly give forensic linguistics a more secure status as a ‘science’. However, although it is often claimed that each human being uses language differently and that this difference can be observed as easily and as surely as a fingerprint, it is, in reality, impossible to compile a collection of markers which would stamp a particular speaker/writer as unique. For the present, therefore, the notion of ‘linguistic fingerprint’ appears essentially flawed and there is little hard evidence to support it. Accordingly, it is better to focus on the distinctive style of a given person, as detected in a set of known and suspected texts within an inquiry. This is something which I will take up further in my section on authorship attribution. Before doing this, however, it is useful to see some of the ways in which forensic linguistics can be of use.

Forensic Linguistics: Some Areas of Application Forensic Phonetics Phonetic techniques are primarily used in the analysis of the voice as applied in criminal investigation. This comprises technical voice comparisons, lay voice recognition, transcription of spoken language, speech signal enhancement, and the authentication of recordings. Forensic phoneticians conduct speaker identifications, resolve disputed content recordings, and transcribe spoken texts. They are also involved in the setting up of so-called voice line-ups or parades in which not eye- but earwitnesses are asked to take part in order to identify a suspect. The typical questions asked in this context are: Was the anonymous caller the same person as the known speaker? Are the two samples from the same dialect/accent? Is the pronunciation of phonemes similar across the known and questioned voices? The fundamental problem with voice line-ups, however, is that despite the fact that, in a threatening situation, we may be capable of storing more features, generally speaking, our memory for voices fades rapidly in comparison to our memory for faces. Voice identification, therefore, needs to be conducted without delay and treated with extreme caution. For the forensic record, spoken texts—be it interviews, oral statements, or interrogations—have to be transcribed into written form, which often causes problems, as some information might go missing or there may be inaccurate relay of the nuances of the oral text (partly due to lack of contextual information and paralinguistic features). In addition, written

96

Chapter Six

discourse differs considerably in mode of expression from spoken discourse which is strongly context-dependent, as discussed below.

Language in Authority and Power Relations In the United States, the Supreme Court in Miranda v. Alabama (1966) set down the requirement that, prior to the arrest or interrogation of a suspect in a crime, that person must be told that they have the right to remain silent, the right to legal counsel, and the right to be told that anything they say can be used in court against them. Instances of the application of this requirement serve well as an illustration of how speech acts performed by police officers may lead to the apparent ‘consensual’ nature of searches, how questioning can be interpreted as coercive, and how the relationship between authority figures and a suspect/defendant is asymmetric. Consider the following examples, discussed in Solan and Tiersma (2005, pp. 35ff) which on the semantic level cannot be interpreted as directives, yet pragmatically speaking, given the authoritarian context, appear precisely as that: Does the trunk open? You don’t mind if we look in your trunk, do you? Why don’t you put your hands behind your back, all right?

The level of coerciveness increases in ‘requests’ such as: Would you mind if I took a look around here? Well, then, you don’t mind if I look around in the car, do you, or would you?

The police usually lack the authority to make promises such as “We’ll go easy on you if you confess”, yet this is implied in their “requests” to comply. The problem is, as Solan and Tiersma (2005, p. 38) point out, that people who are stopped by the police tend to interpret ostensible requests as commands or orders, yet, in contrast, their own indirect wishes to get a lawyer often go unnoticed (for example, “Maybe I should talk to a lawyer”). This problem is further exacerbated due to problems related to the comprehensibility of the Miranda warning and other police language for many suspects, including defendants who may be (semi-)illiterate, speakers of another language, or too young or mentally-challenged to understand their rights to remain silent and seek legal advice. In any case, the asymmetric nature of the relationship between authority figures (the police) and the defendant—who may be disadvantaged

The Role of Forensic Linguistics in Crime Investigation

97

in some way—can result in a text (such as a record of interview, video or audio recording or written statement) which is considerably at variance with what the suspect would have said had he/she been given the opportunity to make a statement in a non-coercive or less threatening environment. This leads to the conclusion that despite the necessity of strong contextual reliance in the interpretation of speech acts, courts may habitually use out-of-context inferences and entailments to reach decisions.

Discrepancies in Police Reports When establishing the accuracy of police reports and alleged suspect statements one has to consider the relationship between the documents exhibited and the events they purport to describe. What is the time frame? When were the incident notes taken? Is there a chronology and accuracy in recalling the events? Too many common features between the statement and the incident notes, coupled with chronological inconsistency and frequent use of characteristically written rather than spoken discourse, may raise suspicion as to authenticity of the police record of an interview or a statement. For this reason, videotaping, recommended by Solan and Tiersma (2005), has been the law for many years in the UK and Australia, yet in the US it is required in only a few states. Police officers typically use so-called “police speak”, which is relatively easy to detect. It is characterised by efficient and compact set phrases, dense wording in an impersonal, official style, with precise renditions of time, place and sequence, as well as precise descriptions of objects, such as weapons. A very revealing expression, otherwise uncommonly used, is I then + verb as in I then threw the weapon into the river. The alteration by the police of a defendant’s utterances, such that they include damaging remarks, is referred to as ‘verballing’. This illegal practice may be done, for instance, in order to match a defendant to a certain racial profile. Racial profiling refers to the use of an individual’s race or ethnicity by law enforcement personnel as a key factor in deciding whether to engage in enforcement, e.g., make a traffic stop or arrest. (For further, detailed discussion of the language of interrogation and statements, see Ollson (2009, pp 100ff).

Cross-Cultural and Cross-Linguistic Differences in Testimony Linguists, and sociolinguists in particular, study differences in varieties and dialects within a given language, and across cultures and languages.

98

Chapter Six

Unfortunately, this cross-cultural linguistic research may not be taken into account by law enforcement authorities passing crucial judgements related to someone’s guilt or innocence. In this relation, Eades (2008), for example, examines the social consequences of courtroom talk through detailed investigation of the cross-examination of three Australian Aboriginal boys in the case against six police officers charged with their abduction. In her study of Australian courtroom discourse, she discovered that yes/no questions are not considered coercive in Australian Aboriginal interactions, but rather are understood as an invitation to explain or elaborate. Further, the difference in cultural meaning attached to silence can also impact judgements in the courtroom: whereas silences longer than a few seconds are hardly tolerated in Western English-speaking societies, Eades’ courtroom data reports common Aboriginal silence up to 23 seconds. Tag questions can also be a source of misunderstanding in testimony to be interpreted. Whereas negative tag questions in English require a negative answer to deny an accusation (e.g., “You took the money, didn’t you?” “No, I didn’t.”), tag questions in many other languages, including Spanish and some Asian languages, can be answered either negatively or affirmatively with relatively no alteration in meaning. Another example refers to the incorrect interpretation of auxiliaries in the testimony of Rosa Lopez during the highly publicised trial of O. J. Simpson (an ex-American football star and sports announcer, accused of the murder of his ex-wife and her friend). The interpretation contributed to a more coercivesounding cross-examination in Spanish than in the original English. As one can appreciate, ultimately, a person may be deemed guilty due to cross-cultural differences in utterance interpretation as expressed, for example, in syntax, prosody or even non-verbal signals involved in producing a statement. In the next part of the paper, I would like to focus on one area where forensic linguistics is particularly relevant: that of authorship attribution.

Authorship Attribution Authorship attribution is the science of inferring characteristics of the author from the characteristics of documents produced by that author. The key task is to establish who said or wrote something which is to be used as evidence. Attribution is facilitated by measuring word length average, average number of syllables per word, article/determiner frequency, and type-token ratio (a measure of lexical variety). Furthermore, punctuation

The Role of Forensic Linguistics in Crime Investigation

99

in terms of overall density, syntactic boundaries and the measurement of unique words in a text, contribute to solving the task. Both Chaski (1997, 2001) and Kredens (2000) stress the importance of taking the relative frequency of various syntactic markers into consideration. Generally speaking, it is easier to eliminate someone as the author than pinpoint someone with certainty.

Forensic Text Types A forensic text is any kind of text, a written document or an audio or video recording, which is the subject of police investigation or of criminal procedure. The investigative linguist may be called upon to analyse a variety of documents. The text types may include emergency calls, ransom demands and other threats, such as hate mail, aimed at victimising others. In this case the genuine or false nature of the call has to be determined to detect or eliminate a hoax, for example. The same differentiation applies to suicide notes or letters. Last statements, on the other hand, may throw some light on the guilt or innocence of a convicted person, if a death row inmate decides to utter their last words: Well, I don’t have anything to say. I am just sorry about what I did to Mr. Peters. That’s all.

Death row statements either (explicitly or implicitly) may confirm commitment of a crime, or deny it, leaving an impression of innocence behind. They may also denounce witnesses as dishonest or criticise law enforcement as corrupt.

Text Message Analysis Text messages (or SMS) may be analysed for authorship attribution in cases of crimes where, for example, the perpetrator is suspected of sending text messages from the victim’s phone, purporting to be written by the victim. In this case, the forensic linguist attempts to determine the consistently used stylistic features. Statistical analysis of a specialised language database of thousands of text messages from a corpus sampler may facilitate such analysis. The key question here is how to determine the point at which a style change within the texts became evident (the socalled “cut-off point”). This has to be accompanied by compilation of a sociolinguistic profile of the purported author in terms of gender, age, origin, as well as social, educational, and professional background. It is

Chapter Six

100

also important to link the messages at hand by means of cohesive and coherent devices to specify the order in which they were sent. Consistent or inconsistent dialectal features may include, for example, the use of pronouns (my/myself v me/meself). Crucial stylistic features include formation of clusters of words (e.g., ‘want2go’) and their average length and character (phrases/clauses v single words). Length of texts and word length average, punctuation, spacing, etc. play an important role as well. One should also consider individual words and phrases that can be written in more than one way (e.g. ‘av’, ‘hav’ and ‘ave’ for ‘have’), as well as alternative lexical choices—morphological, alphanumeric, letter replacive, orthographic (homophonic and punctuation-related, lower/upper case), or orthographic/phonic reduction, as in: 4u2 fone

gr8t!

r u goin?

However, one should bear in mind that a person’s style of writing or texting is not always consistent and it may change, for example, due to changes in life circumstances, the text type, or addressee relationship. Moreover, a language feature which occurs in a small sample cannot be treated as a ‘constant’ for variation in larger samples. In addition, mobile phone texts sometimes use mixed styles (cf. Olsson 2009: 57ff). On numerous occasions thorough linguistic analysis of the SMS messages sent from a victim’s phone have led to the capture of the perpetrator due to certain idiosyncratic features, such as spacing, non-contraction of positive verbs, using “I’m”/”Im” or owing to inconsistencies in texting styles (e.g. “cu” vs “cya”, “my” vs “me”, “Im not” vs “aint”). (See Amos, 2008, for an interesting account in this regard.).

Variation in Author Texts There are two types of author variation: within and across texts. The former, so-called intra-author variation, refers to the ways in which one author’s texts differ from each other. This may include variation in vocabulary, depending on genre, text type, fiction v non-fiction, private v public texts. However, one has to take such factors into consideration as time lapse between two communications, possible disguise, change in personal circumstances (e.g., language of trauma), cultural changes that may influence, for example, the texting language), etc. Moreover, all authors exhibit variation in genre, text type, and the like, and that variation in short texts can be extreme. Inter-author variation deals with the ways in which different authors vary from each other due to widely different social

The Role of Forensic Linguistics in Crime Investigation

101

backgrounds, levels of education, geographical origin, different types and levels of occupation/profession, and so on. There exists also the short text stability problem: in short text analysis we usually find high intra-author variation and low inter-author variation if the texts are of the same type.

Scientific Methods of Authorship Detection Authorship methods which focus on linguistic characteristics currently have accuracy rates ranging from 72% to 95%, within the computational paradigm. Chaski (2005) presents a computational, stylometric method which has obtained 95% accuracy and has been successfully used in investigating and adjudicating several crimes involving digital evidence. Computer crime investigations, where it is crucial to determine who actually pressed the key on the keyboard, range from homicide to identity theft and many types of financial crimes. Evidence in these cases can be collected using several methods, such as biometric analysis of the computer user, qualitative analysis of any idiosyncrasies in the language in questioned and known documents, or quantitative, computational stylometric analysis. Naturally, the higher the rate of accuracy, the better, but questions related to the likelihood of the contested documents belonging to another suspect have to be answered as well. Chaski and Chemylinski (2005a) have developed a method for decomposing the data into smaller chunks so that a larger set of variables can be used for the discriminating analysis. Chaski and Chemylinski (2005b) also obtained similar results using these variables with logistic regression, that is part of a category of statistical models called generalized linear models. Logistic regression allows one to predict a discrete outcome, such as group membership, from a set of variables that may be continuous, discrete, dichotomous, or a mix of any of these. Stamatatos (2009), on the other hand, presents recent advances of the automated approaches to attributing authorship, examining their characteristics for both text representation and text classification. The focus is placed on computational requirements and settings rather than on linguistic or literary issues. He also discusses evaluation methodologies and criteria for authorship attribution studies. An important question is how to discriminate between the three basic factors: authorship, genre, and topic. Are there specific stylometric features that can capture only stylistic, and specifically authorial, information? The application of stylometric features to topic-identification tasks has revealed the potential of these features to indicate content

102

Chapter Six

information as well (cf. Clement and Sharp 2003; Mikros and Argiri, 2007). It seems that low-level features like character N-grams (subsequences of n items from a given sequence, for example, phonemes, syllables, letters, or words) can successfully be applied in stylistic text analysis (cf. Keselj et al. 2003; Stamatatos 2006; Grieve 2007). A crucial need is, however, to increase the available benchmark corpora so that they cover many natural languages and text domains. It is also very important for the evaluation corpora to offer control over genre, topic and demographic criteria.

SMS Authorship Attribution In the face of increasing amount of digital evidence available on cellular phones and, consequently, the necessity to detect SMS (text) authors in criminal persecution cases, Mohan, Baggili and Rogers (2010) propose an N-grams based approach for determining the authorship of text messages. The method shows encouraging results in identification of authors. A token is generated by moving a sliding window across a corpus of text where the size of the window depends on the size of the token (N) and its displacement is done in stages, each stage corresponding to either a word or a character. Since SMS messages are normally very brief and lack many syntactic features, in the forensic analysis of these messages there is a need for high processing speed because, frequently, someone’s life may be at stake. An N-gram approach for an SMS corpus seems to find application under such conditions and is said to predict the author with an accuracy of 65-72% when the samples of SMS messages are small and the number of possible authors is comparably large.

Forensic Linguistic Controversies In the final section of this paper, I will turn to some of the controversies remaining for the forensic linguistics, and the future prospect for the science.

Speaker Identification One of the controversies discussed in, for example, Hollien (2001), is the disagreement in the so-called ‘scientific community’ on the degree of accuracy with which examiners can identify speakers under all conditions. Surprisingly, many suspects will voluntarily give a sample of their voice

The Role of Forensic Linguistics in Crime Investigation

103

for comparison purposes. Vocal ‘disguises’, however, can be very difficult for the examiner to deal with and the probability of determination is lower than with normal voice samples. To prevent problems, investigators need to request that the court order specify in detail that the suspect give a sample of his or her voice, repeating the phrases of the questioned call, in a natural conversational voice (or in a similar disguise, if that is the case) and that such sample shall be given at least three times and to the reasonable satisfaction of the investigator. Voice specimens obtained with such specific instructions are usually very satisfactory for comparison purposes. There is presently, however, no universal standard for the number of words required for identification. It does vary from a minimum of 10 for some agencies and 20 for others. According to Hollien (ibid.), spectrographic voice identification assumes that intra-speaker variability (as discussed above) is discernible from inter-speaker variability (differences in the same utterance by different speakers); however, that assumption is not adequately supported by scientific theory and data. Viewpoints on actual error rates are presently based only on various professional judgements and fragmentary experimental results rather than from objective data representative of results in forensic applications.

Testimony Controversies also arise in relation to witness/police testimony. All the cases of second-hand verbal (apparently verbatim) material (cf. “I don’t know exactly what he said, but I know he said he did it” in Solan and Tiersma, 2005: 98) can be considered unreliable since, as discussed below, human memory is incapable of retaining the exact wording even after a couple of seconds, not to speak of months or years. Moreover, reproduced utterances may be presented in isolation, lacking the original paralinguistic and situational (pragmatic) context. There also remains a great deal of research to be done to increase our insight into the effect of estimator variables on speaker identification by ear witnesses. It should for the time be treated with considerable caution. ‘Scientific’ criteria for court admissibility of testimony still pose a problem as they differ from country to country and from state to state (as in the case of the US). Required qualifications of examiners and presenters of forensic linguistic material—so-called ‘forensic experts’—have not yet been clearly specified, either.

104

Chapter Six

Impressionistic Likelihood and Veracity of Statements As already mentioned above, one may question the admissibility of witnesses’ oral evidence and statements, as well as judges’ decisions based on impressionistic linguistic witness evidence (e.g., reliability of memory, statements deprived of context and pragmatic implications, etc.). Veracity refers to truthfulness of a spoken or written testimony. When defendants feel challenged in this respect, they may suddenly become conscious of their pronunciation (or hyper-correct, in sociolinguistic terms). Despite the fact that some witnesses claim that they can remember exact words of a defendant months or even years later, it is doubtful if this is ever accurate. This seems even less likely, when more than one person, for example, a number of police officers, quote a suspect verbatim after a considerable time lapse. Hence a question arises: how long, in reality, can one can remember what someone else has said—word for word? As Clifford and Scott (1978) state, the upper limit for short-term memory is 79 items, beyond which meaning may be retained but not the actual wording. Moreover, an average recall level is about 30-40% already after a few seconds. In addition, the usage of generic language or an incongruous register when a specific register is normally used leads the forensic linguist to raise doubts about the genuineness of a given statement.

Can Forensic Linguistics Establish Guilt or Innocence? By meeting scientific forensic criteria and presenting convincing linguistic evidence in court, forensic linguists can certainly contribute to pronouncing someone innocent. They can also prompt admittance of guilt. Forensic linguists may be asked to investigate recorded police interrogations to decipher whether or not a person knowingly admitted guilt, underwent just interrogation or understood the conversation conducted throughout the interview. Since recorded interviews can be admitted in court as evidence, dialogue analysis may be carried out to (dis)prove guilt and determine potential inconsistencies in the interviewing process, making recordings inadmissible in court. The defence can therefore show that the recorded language does not necessarily indicate the defendant’s guilt.

What Is a Reliable Sample? Author identification is a very interesting and potentially useful area in determining guilt, but it is restricted by the fact that documents in a

The Role of Forensic Linguistics in Crime Investigation

105

forensic setting (ransom notes, black mail, etc.) are usually much too short to make a reliable identification. Moreover, which linguistic features are reliable indicators of authorship, and how reliable those features are, remains to be discovered. As Tiersma (ibid) points out, research is ongoing, and the availability of large corpora of speech and writing samples suggests that the field may advance in the future (although the typically small size of the documents in most criminal cases will always be a problem). It is therefore crucial for the attribution methods to be robust and applicable to a limited amount of short texts. However, several important questions remain open in relation to the authorship attribution, the most important issue being the required text-length. Despite the fact that various studies have reported promising results with short texts (with less than 1,000 words; cf. Sanderson and Guenter, 2006; Hirst and Feguina, 2007), it has not yet been possible to define a text-length threshold for reliable authorship attribution. In the final section of this paper, I want to turn to some of the future challenges for forensic linguistics and possible ways towards scientific legitimisation of the discipline.

Future Prospects of Forensic Linguistics Will forensic linguistics ever become an established discipline, on a par with scientific forensic methods of providing criminal evidence? From the perspective of its international development the following challenges for the discipline emerge, before such a status can be achieved: x x

x x x

the integrated study of forensic linguistics/language and the law across different judicial systems and geographical boundaries; the development of replicable methods of analysis to be used in expert witness evidence in order to ensure internal and external validity in research; extensive detailing of codes of good practice and conduct; cooperation of International Association of Forensic Linguistics (IAFL) with other associations and societies of forensic sciences; certification of forensic linguistics as a scientific discipline, i.e., universal acceptance of linguistic evidence along other forensic evidence (e.g. as fulfilling the Daubert standard in the USA).

106

Chapter Six

It seems that the future of forensic linguistics lies with corpus-driven approaches (cf. Kniffka 2007). The forensic linguistic community also needs to bring together relevant scholars and linguistics experts of nonEnglish backgrounds with those of English-speaking backgrounds. Kniffka (ibid.) implies that the English-speaking work on forensic linguistics has not always been aware of work published in German, or other languages for that matter. Kniffka claims that the forensic linguistic work in Germany was already well advanced when it was only just beginning in English-speaking contexts.

Summary and Conclusion The present paper has offered a brief overview of the interdisciplinary field of forensic linguistics and illustrated some of its applications, such as pragmatic analysis and various scientific methods of authorship attribution, serving the law and law enforcement. The difficult role of linguists in court testimony is discussed by, for example, Solan and Tiersma (2005) who state that, although not always permitted in court in the end, the linguistic evidence may be helpful to law enforcement in investigating a crime or to lawyers preparing for trial. As a matter of caution one may add that testifying linguists should not, however, state conclusions that indicate more than the evidence presented. No matter how strongly the linguist is convinced that the defendant is innocent, he/she should restrict their opinion to only stating the degree of probability of, for example, a confession being ‘verballed’ by police officers. Moreover, forensic linguists need to stay impartial at all times, as they serve the law in the role of experts and cannot under any circumstances side with defence or prosecution. Despite the fact that linguistic expertise has been frequently favourably compared to fingerprint or DNA evidence, the current state of the art in practices such as voice identification and authorship attribution has not yet reached the same level of reliability. At most, linguistic expertise facilitates elimination of a suspect as the perpetrator, but is not in a position to identify one with certainty (cf. Solan and Tiersma, 2005: 242). Yet, advances in technology and science, as demonstrated above, allow experts to compare documents and voice recordings more quickly and more easily than before. Computer assistance, such as the Federal Bureau of Investigation’s Communication Threat Assessment Database (CTAD), makes it possible to break forensic linguistic data into numerous categories and to make rapid assessments. These developments promise continued expansion of role of forensic linguistics.

The Role of Forensic Linguistics in Crime Investigation

107

Bibliography Amos, O. 2008. The text trap. The Northern Echo. Retrieved January 5, 2012 from http://www.thenorthernecho.co.uk/features/leader/207 6811.the_text_trap/ Chaski, C.E. 2005. Empirical evaluations of language-based author identification techniques. International Journal of Speech, Language and the Law, 8 (1), pp. 1-65. —. 2005. Who’s at the keyboard? Authorship attribution in digital evidence investigations. International Journal of Digital Evidence 4 (1), pp. 1-13. Chaski, C. E., and H. J. Chmelynski. 2005a (pending publication). Testing twenty variables for author attribution by discriminant function analysis. Chaski, C. E., and H. J. Chmelynski. 2005b (pending publication). Testing twenty variables for author attribution by logistic regression. Clement, R., and D. Sharp. 2003. N-gram and Bayesian classification of documents for topic and authorship. Literary and Linguistic Computing, 18 (4), 423-447. Clifford, B.R. 2009. The role of the expert witness. In G. Davies, R. Bull and C. Hollin (eds.). Forensic Psychology. New York: Wiley. Eades, D. 2008. Courtroom talk and neocolonial control. Berlin and New York: Mouton de Gruyter. —. 2000. I don’t think it’s an answer to the question: Silencing aboriginal witnesses in court. Language in Society, 2000 (29), pp. 161-195. Grieve, J. 2007. Quantitative authorship attribution: An evaluation of techniques. Literary and Linguistic Computing, 22 (3), pp. 251-270. Hirst, G. and O. Feiguina. 2007. Bigrams of syntactic labels for authorship discrimination of short texts. Literary and Linguistic Computing, 22 (4), pp. 405-417. Hollien, H. 2001. Forensic Voice Identification. London: Academic Press. Keselj, V., F. Peng, N. Cercone, and C. Thomas. 2003. N-gram-based author profiles for authorship attribution. Proceedings of the Pacific Association for Computational Linguistics, pp. 255-264. Kniffka, H. 2007. Working in Language and Law: A German Perspective, Basingstoke: Palgrave Macmillan. Kredens, K. 2000. Forensic linguistics and the status of linguistic evidence in the legal setting. Unpublished Ph.D. dissertation. University of Lódz. Leech, G. 1983. Principles of Pragmatics. London: Longman.

108

Chapter Six

Mikros, G. and E. Argiri. 2007. Investigating topic influence in authorship attribution. Proceedings of the International Workshop on Plagiarism Analysis, Authorship Identification, and Near-Duplicate Detection, pp. 29-35. Olsson, J. 2009. Word Crime: Solving Crime through Forensic Linguistics. New York and London: Continuum International Publishing Group. —. 2008. Forensic Linguistics. New York and London: Continuum International Publishing Group. Sanderson, C. and S. Guenter. 2006. Short text authorship attribution via sequence kernels, Markov chains and author unmasking: An investigation. Proceedings of the International Conference on Empirical Methods in Natural Language Engineering, pp. 482-491. Morristown, NJ: Association for Computational Linguistics. Solan, L.M. and P.M. Tiersma. 2005. Speaking of Crime: The Language of Criminal Justice, Chicago and London: Chicago: The University of Chicago Press. Stamatatos, E. 2009. A survey of modern authorship attribution methods. Journal of the American Society for Information Science and Technology, Volume 60, Issue 3, pp. 538–556. —. 2006. Ensemble-based author identification using character n-grams. Proceedings of the 3rd International Workshop on Text-Based Information Retrieval, (TIR’06), pp. 41-46. Svartvik, J. 1968. The Evans Statements: A Case for Forensic Linguistics. Gothenburg Studies in English, 20. Tiersma, P. M. 2009. What is language and law? And does anyone care? In F. Lorz, A. and D. Stein, (eds.) Law and Language: Theory and Society. Loyola-LA Legal Studies Paper No. 2009-11.

CHAPTER SEVEN UNIVERSITY ENGLISH STUDIES IN MULTILINGUAL CONTEXTS: WHAT ARE THE PROSPECTS? JAMES A. MOODY

Abstract When university English studies are examined in the context of multilingual societies today, there is often a perception that standards are declining even as uses of the language are expanding. This apparent paradox is attributed to a pedagogical failure to consider the relationship between English and other languages and its position as an element in an overall communicative system. Evidence from one speech community, the Sultanate of Oman, serves as an example of the social and linguistic forces contributing to an increasingly complex language ecology. Some common misconceptions concerning the nature of English studies in such circumstances are dealt with, and a suggestion is made for how they might be overcome through a degree programme on language in the workplace. This approach would help English departments meet their responsibilities by improving students’ communicative skills, developing local scholarship and responding to national interests.

Keywords: multilingualism, context, language change, English language teaching (ELT), English for specific purposes (ESP), language in the workplace

English Standards and Language Change In many multilingual communities where English is a non-native language, its forms, functions and status are changing at the beginning of

110

Chapter Seven

the twenty-first century. This poses a challenge for university English departments. On the one hand, although English language teaching (ELT) has received considerable support, students' abilities have been found inadequate. In Anglophone Africa, for example, “[i]n spite of extensive spread of English to earliest levels of education and…tremendous resources invested in its promotion, there have been numerous claims of ‘falling standards’…” (Mazrui, 2004, p. 41). In Asia, similar criticism has been made in Brunei (Wood, et al, 2008), and in Singapore, the negative influence of "Singlish" on English has been of concern in the national media (Miller, 2010). With regard to the Arab world, a 2003 United Nations report concludes that “facility with the English language is waning” (cited in Karmani, 2005, p. 94). On the other hand, demands for English are growing, and it continues to be a major means of communication not only globally but within these communities as well. Crystal (2003, p. 6) has estimated that one-quarter of the world’s population is “fluent or competent” in English. The only language approaching its numerical hegemony is Mandarin Chinese (Graddol, 2006, p. 62), but as a medium for worldwide communication English has few rivals. The pervasiveness of English in multilingual communities is evident in such ordinary aspects of life as informal conversation, the media and commercial signs in cities of Africa, Asia and the Middle East, where English mixes with other languages in daily communication. Consequently, ideas about the nature of the language itself are changing because of its uses in particular social contexts (Higgins, 2009; Pennycook, 2010). The study of English as a non-native language at universities in such environments should be considered against this background. This chapter argues that in order to improve students’ abilities to communicate in English, university departments need to give more cognizance to the changing relations between language and society. It will first address factors contributing to the apparent paradox that, even as the English language is expanding throughout the world, criticism of its use is also increasing. The purposes English serves for native speakers of other languages and its integration into the communicative systems of multilingual communities will then be explored. Next, attention will be given to possible reasons for shortcomings in ELT, which involve misconceptions about links between language education and linguistic reality. Finally, a possible means of resolving this paradox will be proposed, one which entails developing an aspect of English studies previously deemed inappropriate to university education. A suggestion will be made for how a university degree concentrating on language in the

University English Studies in Multilingual Contexts

111

workplace might respond to social conditions and national demands while at the same time satisfy academic criteria and improve communicative competence.

The Position of English in Multilingual Communities One result of the spread of English is that it is no longer a firstlanguage for most of its speakers. Crystal (2003, p. 69) puts the global ratio of native to non-native users as one to three. Consequently, the main reason for using English worldwide is for communication between nonnative speakers. This fact has consequences for its forms and functions as well as its pedagogy. Kachru’s (1985) familiar model of inner (first language), outer (second language) and expanding (foreign language) circles of communities of English users obscures contemporary sociolinguistic complexities (Bruthiaux, 2003). The inner circle has lost its authoritative status, and the distinction between outer and expanding circles is increasingly blurred. In ELT, this has led to what Graddol (2006, p. 87) refers to as a “radical and controversial” response, English as a lingua franca: “the study of how non-native speakers use English among themselves”. As a result of these changes, terms such as standards, competence and fluency come to be focused less on linguistic systems than on judgments of speakers. This is evident in the results of Higgins’ (2003, p. 640) research into language attitudes: people who speak English in the outer circle are confident in their use even when it does not conform to judgments about correctness in the inner circle. In this light, standards of native-speaker competence, for long considered a goal of ELT, are now problematic. For instance, Canagarajah (1999) holds that “standard English” worldwide is developing into a “hybridized and diversified” language, so that a competent speaker is one who can successfully “arbitrate between different cultures, communities, and discourses” (p. 181). Effective communication, then, becomes the primary criterion in determining acceptability. In fact, native speakers in the inner circle are losing their privileged position in that many of them are weak in communicating with non-native speakers (Graddol, 2006, p. 87). Since most English speakers also use other languages, monolingualism is no longer the norm; yet among ELT practitioners, multilingualism is frequently stigmatized (Karmani, 2005, pp. 93-94). To underscore this inconsistency, Phillipson and Skutnabb-Kangas (1996) have developed a model of contrasting approaches to language policy for multilingual communities. At one end of the continuum is the diffusion-of-English

112

Chapter Seven

paradigm, embodying the assumption that the homogenisation resulting from globalisation of English is the key to modernisation and prosperity. One result of this policy is that educational resources are concentrated on English, with the consequence that other languages are neglected. At the other extreme is the ecology-of-language paradigm, which acknowledges interaction between languages, their mutual dependence and connections to other aspects of the social and geographical environment. Thus, support is given to minority languages, and equality among speakers of different languages is promoted. In one sense, the spread of English has effected a movement in the ecology-of-English direction, as previously monolingual communities have grown more tolerant of linguistic diversity. But under multilingual conditions, extreme tolerance of a powerful global language can result in its domination through the diffusion-of-English model. English as an agent of neo-imperialism has been a theme in considerations of the spread of English in many parts of the world (Pennycook, 1994; Mazrui, 2004; Karmani, 2005). Phillipson and Skutnabb-Kangas (1996) refer to the continuing strength of the “myth of monolingualism” (p. 438) and its erroneous implication that language contact inevitably leads to conflict. Thus, they caution that “English can serve many useful purposes but will do so only if the linguistic human rights of speakers of other languages are respected” (p. 447). People who have English as a non-native language in multilingual communities often show mixed attitudes towards it. English may bring economic and social advantages, but it can also threaten traditional values and culture. However, as the language develops as a lingua franca and is used along with other languages, the consequences are often less fear of domination than pragmatic acceptance. The example of opinions of university students in Saudi Arabia who were investigated by Al Haq and Smadi (1996) seem typical: they are unconcerned about threats from English, which they see as a necessary tool for modernisation and advancement as well as a means for social enrichment and even for the spread of Islam among non-Arabic speakers. Under multilingual conditions, languages tend to lose their identification with particular cultures. This is because, as Pennycook (2007a) argues, languages are less autonomous entities than responses to unique social requirements, which Anchimbe (2007, p. 3) connects to migration and urbanization. In response to such contingencies, multilingual groups create unique "modes of identity" (Pennycook, 2007b, p. 4) since they "appropriate [language] materials differently" (p. 6). In many contexts English is not simply a lingua franca; it also functions as an element in what can be

University English Studies in Multilingual Contexts

113

termed a superordinate system, one in which all languages in use in a given milieu are mutually dependent. Evidence for this can be seen, for instance, in the phenomenon of code-switching, which characterises multilingual communities in Africa (Moody, 2001), the Middle East (Sallo, 1992) and elsewhere. The mixing of mutually intelligible languages increases meaning potential, so that speakers are able to communicate more effectively than would be possible within a single language. It is significant that such concerns rarely figure in ELT in such societies.

Influences on University English Studies External and internal factors affect English departments at universities in multilingual communities. Some of these influences obscure the nature of multilingualism as outlined above. The challenge is to move towards the ecology-of-language paradigm by identifying potential tensions and using them as incentives for change.

External Factors In many universities, supplying secondary school teachers is becoming less of a priority now than in the past, as national graduates fill available positions in schools due to localisation policies. However, there are still reasons for English departments to educate future teachers because of normal attrition rates and increases in school-age populations. In addition, as universities seek international accreditation and influence, they may wish to serve regional, in addition to national, stakeholders. Another consideration is that manpower requirements should be met not simply in terms of quantity but also with regard to the quality of graduate teachers, possibly through in-service courses and post-graduate certificates to improve secondary teaching. But to justify a degree programme on the grounds of providing teachers is a circular argument. It should first be asked why English teachers are needed at all. An adequate answer requires attention to the roles English plays in society, where the justification for teaching and studying language lies. To claim that university students will use what they learn to teach English obscures this fundamental point. The fact that there is a declining market for graduate teachers in secondary schools could encourage a reconsideration of both academic programs and their rationale, in broader social terms. Phillipson and Skutnabb-Kangas (1996, p. 439) note that the ELT profession itself evinces the diffusion-of-English paradigm. Pressures are

114

Chapter Seven

exerted by institutions, governments, and commercial interests. The indiscriminate adoption of one-size-fits-all English-language textbooks and international examinations, for example, has been frequently criticised for neglecting local requirements (see Moody, 2009, pp. 106-108 and references therein). In particular, distinctions are not always made between English for non-native speakers in predominantly first-language English environments and English for non-native speakers in multilingual contexts. The authority of international publishers and examining boards, which continue to emphasise native-speaker standards, may be difficult to resist when resources required to develop alternatives are limited. Nevertheless, closer attention needs to be given to local uses of English in materials, methodology, and assessment.

Internal Factors The failure of university English studies to respond satisfactorily to local conditions can also be attributed to four prevailing factors within English departments: the fragmentation of courses, an emphasis on theory and knowledge over practice, misconceptions about students’ motivation and assumptions of their inadequacy. English departments in multilingual communities typically comprise distinct academic disciplines: language, linguistics, literature, and, sometimes, translation. Such an umbrella structure could be a stimulus for integrated courses and programmes centring on community interests. But this occurs rarely; instead, disciplinary boundaries tend to be strictly preserved, which results in a tendency to look inward rather than outward. Consequently, from students’ perspectives, the courses they take can be unconnected with each other or with wider social issues. For example, K. M. Al Balushi’s (2010) recommendation that university English studies in Oman be organised on the basis of discourse, how language is employed in social action, is motivated by his perception of the present fragmentation and lack of focus. Related to this is an implicit assumption in academia that the more advanced and specialised the study, the more central is theory and knowledge over practice. The term skill often has negative connotations in university ELT, and teaching the use of English can be viewed as subsidiary and remedial. But since perfection in language is impossible, strengthening skills is a life-long pursuit. When English departments neglect this fact, the result is often, as Haggen (1999) says, “…that too many students graduate with a knowledge of transformational grammar

University English Studies in Multilingual Contexts

115

and Shakespeare’s plays but without having attained the ability to produce well-constructed, error-free sentences...” (p. 33). Motivation in non-native language education has long been considered in terms of an integrative-instrumental dichotomy. Whereas the former depends on individual learners’ positive attitudes to native speakers, the latter is based socially on their desire to achieve practical goals. Instrumental motivation can be taken for granted in most multilingual contexts insofar as people learn English for economic reasons. It has been shown that students in Saudi Arabia (Al Haq and Smadi, 1996) and Japan (Norris-Holt, 2001) want to learn English even when they do not admire native speakers. However, in much of the material, methodology, and assessment of English at tertiary levels in multilingual communities, integrative motivation is tacitly assumed (Moody, 2009, p. 103). Under these circumstances—when courses are fragmented, when knowledge and theory take priority over skills and practice, and when motivation is misunderstood—there can be “an inbuilt ideological positioning of the student as outsider and failure” (Graddol, 2006, p. 83). In this way, the diffusion-of-English paradigm is perpetuated through privileging native-speaker standards. Some university graduates may indeed go on to live, work or study in inner-circle environments, but it can hardly be claimed that the central responsibility of local universities is to train them for this purpose. What is required is an approach to English studies that builds positively on students’ actual language experience and capacities.

Multilingual Reality: the Case of Oman The complex language ecology within which English departments operate and which comprises the linguistic background of university students can be illustrated by a specific but in some ways typical multilingual community, the Sultanate of Oman. The geographical position of Oman at the convergence of the Arabian Gulf and the Arabian Sea has meant that it has for several centuries been a crossroads of commercial, cultural and linguistic exchange among Arabia, Africa, and the Indian subcontinent. As a result, coastal Oman is traditionally multilingual. K. Al Busaidi (1995, pp. 83ff) provides a thorough account of the rich mix of languages in Oman and the reasons for it. In addition to Arabic, the most common first-language, several other native languages are spoken. For example, only 75% of Al Busaidi’s representative selection of respondents spoke Arabic before entering school (p. 221). In addition, the number of non-indigenous languages

116

Chapter Seven

increased during the last quarter of the twentieth century, when Oman, along with other Arabian Gulf states, experienced a major influx of foreign workers, mainly from the Indian subcontinent (see, for instance, Karmani, 2005, p. 90). Today, English is widely used in Oman between Arabic and nonArabic speakers (S. B. Al Busaidi, 1998, p. 287 and passim), but it often varies from native English of the inner circle (K. Al Busaidi, 1995, p. 301). Mainly because of urban multilingualism, ‘nonstandard’ forms of several languages have developed throughout the Gulf. The linguistic complexity of these societies is suggested by Smart’s (1990, p. 83) summary (slightly revised here) of typical communicative patterns: Arab > < Arab (either standard Arabic or vernacular dialects) Arab> < European (English, possibly “nonstandard”) European > < European (English) Indian, professionals > < Indian manual labourers (Hindi/Urdu or possibly “nonstandard” English) Indian, especially manual labourers> < Arab (“nonstandard” Arabic or Gulf Pidgin or “nonstandard” English).

Since the 1970s, when a significant number of English-speaking Omanis, whose ancestors had left Oman, “returned” from overseas (K. Al Busaidi, 1995, p. 131), English has been, along with Arabic, a language of commerce, bureaucracy, education, and the media. In addition, it is becoming a vernacular in informal encounters and, like Arabic, is developing "nonstandard" forms. In addition, speakers are adopting communicative patterns in which English combines with other languages (R. Al Balushi, 2010). Thus, English does not function in isolation but as one among several options. In these circumstances, the selection of English can itself have communicative significance. For example, codeswitching into and out of English may signal a particular illocutionary force, or assert a relationship, or mark a rhetorical point, as Al Balushi suggests. Hence, English in Oman is not simply an autonomous lingua franca but is also interacting with and dependent on other languages, as one element in the language ecology.

Misconceptions about English Studies in Multilingual Communities How can university English departments respond positively to such linguistic diversity? Many of the observations about falling standards spring from an acceptance of the diffusion-of-English paradigm and a

University English Studies in Multilingual Contexts

117

neglect of the position of English in the overall ecology of language. For example, Edward Said (1993, p. 370) reports on a visit to an unnamed Arabian Gulf country in 1985 to evaluate university English studies. He regrets the fact that most students were learning English for occupational reasons and cautions that under these circumstances the language is “terminally assigned” to a technical lingua franca and sinks to “a low, uninteresting, and attenuated level”. The literary scholar George Steiner has similar misgivings about the "twofold enervation" that results from the "internationalization" of English; in his opinion, it “impoverishes” (1975, p. 470) both local languages and English. However, to declare that a language is uninteresting or attenuated is to condemn the reasons why people use it and ultimately to denigrate users themselves. A lingua franca may diverge from inner circle norms, but it can nevertheless be used creatively in non-native speaker interaction (Canagarajah, 2006). There are complexities, subtleties, and critical standards for all varieties and combinations of languages used communicatively. What is required is a recognition of the meaningpotential of any linguistic system actually employed and the ways in which people “construct the world through language” (Pennycook, 2010, p. 108). Evidence of the creative possibilities of mixed superordinate codes, consisting of several languages, and the processes by which they arise in southern Africa, for example, is provided by Chisanga and Kamwangamalu (1997) and Moody (2001). Attitudes such as Said’s and Steiner's, then, embody misconceptions about reasons why most people learn non-native languages, the nature of language change and language pedagogy. When, in multilingual environments, the forms of English—its phonology, syntax, and pragmatics—depart from inner circle norms, language learning becomes a matter of developing a repertoire of communicative contexts and engagement in “a process of semiotic reconstruction” (Pennycook, 2007b, p. 115). A rejection of the idea that a language can be successfully studied apart from its social uses prompts Long to propose, “Every language course should be considered a course for specific purposes” (2005, p. 1). Most people who learn English as a non-native language have specific reasons for doing so. If university students seem uninterested, this may be because they do not see clearly how the English they study is or will be valuable to them outside the classroom. Specificity and purpose are concepts about which teachers and students may differ. For example, when students at colleges in Oman were asked why they were studying English, 76% reported their highest priority to be to use it in the workplace; this was in contrast to their teachers, 81% of

118

Chapter Seven

whom said the main reason was to help students academically (Al Husseini, 2004, p. 150). Similar discrepancies have been noted in Jordan (Zughoul and Hussein, 1985). Such cases indicate that resolving the apparent paradox of declining abilities in the midst of language expansion, will involve the mediation of English studies through students’ multilingual worlds. One means of achieving this will now be explored.

Resolving the Paradox The area of ELT known as English for Specific Purposes (ESP) assumes that the more precisely communicative tasks are defined, the more efficiently they can be accomplished. But ESP has been regarded as marginal to university study. Its concerns are usually limited to training in basic skills for academic work (EAP) and for occupations (EOP). In an early account, Kennedy and Bolitho (1984, p. 137) point out that ESP can be boring when there is a mechanical emphasis on an end product rather than on processes involved in achieving it. They insist, though, that ESP need not be uninspiring or restrictive. Sophisticated communicative skills are important across a range of purposes and are far from being formulaic or predictable. For example, ESP can offer multilingual students practical opportunities to assess communicative situations and select from various language options. The importance of communication in the workplace has long been recognised. Professional expertise required in job performance is inadequate if employees do not know how to use language effectively (Holmes, 1995, p. 344). Much of the pedagogy and scholarship on workplace English, however, has been dominated by the diffusion-ofEnglish paradigm. Assumptions have been made from the perspective of the inner-circle, and there is a tendency to assume that principles and practices hold globally. When multilingual communication has been taken into account, it is commonly presented as a hurdle to be overcome, a problem in understanding between native and non-native speakers (for example, in a popular textbook by Adler and Elmhorst, 2008). Coleman (1989, p. 8) remarks that research in this field too has tended to be restricted geographically to Anglo Saxon societies. But the need for localized ESP has been felt in many multilingual communities. In Oman, K. Al Busaidi (1995, p. 287) has claimed that there is “overwhelming” support for both English and Arabic in the workplace. The Professional Development Manager of the Abu Dhabi Educational Council has recently echoed this need throughout the Arabian Gulf: “The expansion of specialized economic cities and the increased

University English Studies in Multilingual Contexts

119

number of e-government initiatives … demands improving English in the workplace” (“British Council…”, 2007, para. 4). The value of EAP and EOP in the wider Arab world has been recognised at least since the publication of the book English for Specific Purposes in the Arab World by Swales and Mustapha in1984. But much of the work since then has concentrated on basic skills at vocational and non-professional levels (see Al Husseini, 2004, for example). Therefore, it is proposed that universities in multilingual communities offer a degree in language in the workplace. The identity of English as a lingua franca and as an element in a superordinate system would be addressed by focusing on the domain of work and asking why specific forms of communication are necessary or advantageous in this context. Specifically, a series of concurrent and interrelated core courses would be offered in three areas: communication theory, practical skills, and the local history and sociolinguistics of language use. The skills courses would concentrate on strategies to deal with matters such as conflict resolution and worker-customer/client interaction. Students would apply insights from sociolinguistics to existing multilingual conditions. Theoretical considerations are required for developing productive skills since “understanding of the wider context is crucial for…discourse at a local level” (Holmes, 1995, p. 348). The programme would be interdisciplinary, including the study of other languages spoken in the community, and there would be a required research essay on a topic involving multilingual workplace communication. A semester of work-attachment during the final year would assign students to an organisation in the public or private sector. They would write a report on this experience, analysing successes and failures in terms of the concepts they have studied. Arab university students often overestimate their abilities in English (Zughoul and Hussein, 1985, p. 148) and fail to engage in critical self assessment until employment after graduation. A period of work-attachment would enable them to develop insights into their own communicative capacities in an actual work environment and to take steps to improve prior to employment. The justification for this type of study rests on three grounds: its viability as a field of scholarship, its potential in encouraging students to study English and its contribution to national development. Multilingual communities offer opportunities for theoretically driven research that merges scholarship in sociolinguistics, pragmatics and English among non-native speakers. In addition, the programme would provide a unique perspective on workplace studies by examining English in the context of multilingualism. It would focus students’ academic work and respond to

120

Chapter Seven

their instrumental and economic motivation. By placing skills within a theoretical framework, it would help to resolve conflicts between teaching about English and strengthening the ability to use it. Finally, the programme would contribute to meeting national needs for improved communication by fostering relevant skills and insights into why they are necessary in terms of the local language ecology.

Responsibility of English Departments There are degrees of appropriateness and standards for judgment in language use found, not in textbooks, but in groups of speakers. If English departments neglect this fact, they will be isolated from their surroundings and in danger of assessing validity according to criteria of scholarship and teaching developed elsewhere. Academics in multilingual contexts should not merely apply existing ideas and models from the inner circle to local data but should also be engaged in formulating, arguing and testing new theories. What, for example, can the world learn from the complex linguistic past and present of countries like Oman? What can be discovered about principles of interaction between languages from evidence provided by exchanges in Arabic and English in cities of the Arabian Gulf? According to various communication theories, meaning is achieved by relating new information to present knowledge, through pre-existing mental patterns, and the contexts in which discourse is experienced become ‘semantic constructs’ that permit the creation of meaning (Widdowson, 2004, p. 54). A pretextual motive, then, is crucial, and “[w]hat is relevant…is what users choose to make relevant in relation to what they are processing language for” (p. 76). This principle applies to the work of university English departments in multilingual communities. If ELT is unconnected to students’ experience, education may, as a result, be literally meaningless. Mental patterns will not be activated; there will be a lack of context or pretext through which meaning can be created since there will be few means of processing the raw material received. This paper has suggested one way of providing such a means. A programme in language in the workplace would initially focus on what students already experience (the language ecology of their community and the role of English within it). Approaching English studies in this way entails a new kind of pedagogy, one in which a teacher is not the source of information to be mastered by a passive student, but rather one in which teacher and student are engaged cooperatively. There are aspects of communication in multilingual communities about which students may be

University English Studies in Multilingual Contexts

121

better informed than teachers (especially if the latter are expatriates). The teacher’s role in these circumstances is to first make conscious and articulate this experience and then to subject it to critical and analytical scrutiny. Academic empowerment, for students as well as for scholars, results from considering whether and how theoretical approaches developed elsewhere apply to local conditions. Writing specifically about Oman, Al Kharusi (2010, p. 6) is concerned with a “sense of alienation among the young”—a perception that has been borne out in the wider Middle East by events of the Arab spring of 2011 in Egypt and elsewhere. The warning that such developments contain for university English studies is that unless cognizance is taken of the linguistic reality of local communities and the increasing ‘hybridity’ of language (Canagarajah, 1999; Higgins, 2009; Pennycook, 2010) within their communicative systems, students will continue to perceive English as unrelated to their particular lives and experience. One way in which the “active engagement” and “social responsibility” that Al Kharusi urges can be accomplished is by focusing on local uses of language through a programme such as the one suggested here for language in the workplace. Al Kharusi concludes, “We should never underestimate the enthusiasm of the young and old once a sense of purpose is awakened in them”. It is a cliché but nonetheless true, although sometimes forgotten: students learn best when they are interested in what they are doing, and this interest in turn must be initiated by what is familiar.

Bibliography Adler, R.B. and J.M.Elmhorst. 2008. Communicating at Work, 9th ed. Boston: McGraw Hill. Al Balushi, K.M. 2010. Teaching English as Discourse in Sultan Qaboos University. Learning and Teaching in Higher Education: Gulf Perspectives 7 (1).Retrieved March 1, 2010 from http://lthe.zu.ac.ae. Al Balushi, R. 2010 (June 27). Majan College R&D Initiative Earns Praise. Times of Oman. Retrieved September 22, 2010 from www.timesofoman.com/echoice.asp?detail=37816. Al Busaidi, K. 1995. English in the Labour Market in Multilingual Oman with Special Reference to Omani Employees. University of Exeter: PhD thesis. Al Busaidi, S.B. 1998. An Investigation into Teachers' and Students' Attitudes towards the Use of the Mother Tongue in the English Foreign Language Classrooms in Oman. University of Reading, U.K.: Masters dissertation.

122

Chapter Seven

Al Haq, F.A. and O.Smadi.1996. Spread of English and Westernization in Saudi Arabia. World Englishes 15 (3), pp. 307-317. Al Husseini, S.S. 2004. An Analysis of the English Needs of Omani Students on Vocational and Technical Courses with Implications for the Design of Foundation Year English Programs. School of Education, University of Leeds: PhD dissertation. Al Kharusi, N.S. 2010 (January 6). Images of solidarity. Times of Oman. Anchimbe, E.A. 2007. Introduction: Multilingualism, Postcolonialism, and Linguistic Identity: Towards a New Vision of Postcolonial Spaces. In E.A. Anchimbe ed. Linguistic Identity in Postcolonial Multilingual Spaces, pp. 1-22. Newcastle: Cambridge Scholars Publishing. British Council Launches English in the Workplace Initiative for the Middle East. 2009. Manamah: British Council. Retrieved January 15, 2010, from http.//www.ameinfo.com/141975.html. Bruthiaux, P. 2003.Squaring the Circles: Issues in Modeling English Worldwide. International Journal of Applied Linguistics 13 (2), pp. 159-177. Canagarajah, A.S. 1999. Resisting Linguistic Imperialism. Oxford: Oxford University Press. —. 2006. Negotiating the Local English as a Lingua Franca. Annual Review of Applied Linguistics 26, pp. 197-218. Chisanga, T. and N M.Kamwangamalu. 1997. Owning the mother Tongue: The English Language in South Africa. Journal of Multilingual and Multicultural Development 18, pp. 89-99. Coleman, H. 1989.Working with Language: An Introduction. In H. Coleman ed. Working with Language: a Multidisciplinary Consideration of Language in Work Contexts, pp. 1-23. New York: Mouton de Gruyter. Crystal, D. 2003. English as a Global Language, 2nd.ed. Cambridge: Cambridge University Press. Graddol, D. 2006. English Next. London: British Council. Haggan, M. 1999. A Linguist’s View: the English Department Revisited. English Teaching Forum 37 (1), pp. 22-29. Higgins, C. 2003. 'Ownership' of English in the Outer Circle: an Alternative to the NS-NNS Dichotomy. TESOL Quarterly 37 (4), 615644. —. 2009. English as a Local Language. Bristol: Multilingual Matters. Holmes, J. 1995. When Small Talk is a Big Deal: Sociolinguistic Challenges in the Workplace. In M.H. Long ed. Second Language Needs Analysis, pp. 344-371. Cambridge: Cambridge University Press.

University English Studies in Multilingual Contexts

123

Kachru, B.B. 1985. Standards, Codification and Sociolinguistic Realism: The English Language in the Outer Circle. In R. Quirk and H.G. Widdowson eds. English in the World: Teaching and Learning the Language and Literature, pp. 11-34. Cambridge: Cambridge University Press. Karmani, S. 2005. Petro-linguistics: the Emerging Nexus between Oil, English, and Islam. Journal of Language, Identity, and Education 4 (2), pp. 87-102. Kennedy, C. and R. Bolitho. 1984. English for Specific Purposes, London: Macmillan. Long, M.H. 2005. Overview: Rationale for Needs Analysis and Needs Analysis Research. In M.H. Long ed. Second Language Needs Analysis, pp. 1-16 Cambridge: Cambridge University Press. Mazrui, A.M. 2004, English in Africa after the Cold War. Clevedon: Multilingual Matters. Miller, S.L. 2010. April 12. Are English Standards Slipping? Straits Times. Retrieved September 5, 2010 from www.straitstimes.com/BreakingNews/Story/STIstory_51344.html Moody, J. 2001. Zambians Talking: Twenty-five English Conversations. Lusaka: The University of Zambia Press. Moody, J. 2009. A Neglected Aspect of ELT in the Arabian Gulf: Who is Communication between? In L.J. Zhang, R. Ruddy, and L. Alsagoff eds. English and Literature-in-English in a Globalized World, pp. 99119. Singapore: National Institute of Education, Nanyang Technological University. Norris-Holt, J. 2001. Motivation as a Contributing Factor in Second Language Acquisition. The Internet TESL Journal 7 (6). Retrieved July 8, 2008 from http://iteslj.org/Articles/Norris-Motivation.html. Pennycook, A. 1994. The Cultural Politics of English as an International Language. London: Longman. —. 2007a. The Myth of English as an International Language. In S. Makoni and A. Pennycook eds. Disinventing and Reconstituting Languages, pp. 90-115. Clevedon: Multilingual Matters. —. 2007b, Global English and Transcultural Flows. London and New York: Routledge. —. 2010, Language as a Local Practice. London: Routledge. Phillipson, R. & T.Skutnabb-Kangas, 1996, English Only Worldwide or Language Ecology? TESOL Quarterly 30 (1), pp. 429-452. Said, E. 1993. Culture and Imperialism. London: Chatto and Windus. Sallo, I.K. 1992. Arabic-English Code-switching at the University: a Sociolinguistic Study. In R.de Beaugrande, A. Shunnaq, and M.H.

124

Chapter Seven

Heliel eds. Language, Discourse, and translation in the West and the Middle East, pp.115-131. Amsterdam/Philadelphia: John Benjamins. Smart, J. 1990. Pidginisation in Gulf Arabic: a First Report. Anthropological Linguistics 32, 88-119. Steiner, G. 1975. After Babel. London: Oxford University Press. Swales, J. and Mustafa, H. eds. 1984. English for Specific Purposes in the Arab World. Birmingham, U.K.: University of Aston (The Language Studies Unit). Widdowson, H.G. 2004. Text, Context and Pretext: Critical Issues in Discourse Analysis. Oxford: Blackwell. Wood, A., A. Henry, A. Clynes, and M.A.S.M.H. Abdullah. 2008. English in Brunei: 'She speaks Excellent English' – 'No he doesn't'. unpublished paper presented at the 13th International Conference on English in Southeast Asia, Singapore: National Institute of Education, Nanyang Technological University. Zughoul, M.R. and R.F.Hussein. 1985. English for Education in the Arab World: a Case Study of Needs Analysis at Yarmouk University. The ESP Journal 4 (2). pp 133-152.

SECTION II LANGUAGES CONSIDERED

CHAPTER EIGHT HOW ENGLISH GRAMMAR HAS BEEN CHANGING GEOFFREY LEECH

Abstract The evidence of electronic corpora shows that English grammar has recently been changing through the increased frequency of some forms, and the decreasing frequency of others. In this way, the boundaries of usage of some constructions (e.g. the present progressive) have been ‘stretched’, and some forms which used to be rather marginal to grammar (e.g., the semi-modals be going to, need to and have to) have been becoming more central. On the other hand, other constructions, such as those using the modal auxiliaries and the passive, have been slowly declining: for example, the modals may, must, need(n’t), ought to, and shall, are much less common than they were fifty years ago. Although the implications of these changes for language teaching should not be exaggerated, it is clear that in ELT less attention should be paid to the declining forms, and more attention should be paid to those that are increasing.

Keywords: English grammar, recent change, frequency, corpora, modals, semi-modals, progressive, passive, grammaticalisation, colloquialisation

Introduction This paper is concerned with the gradual changes that have been taking place in English grammar. The grammar of a language changes slowly and imperceptibly. On the one hand, elements which are normal—or central— in the language gradually become less usual—or marginal—and perhaps will eventually disappear. On the other hand, elements which are rare in

How English Grammar has been Changing

127

the language may gradually become more frequent and more ordinary. Sometimes the boundaries of what is possible in the language become stretched, to take in new possibilities. Native speakers of a language are usually unaware of changes happening in the grammar of their language. If they think grammar is changing at all, they generally imagine that the language is deteriorating, and that standards of ‘good grammar’ are declining. This is not the point of view that will be taken in this paper: it is more helpful to think of grammar adapting to new influences and new possibilities of communication, and the point of view I will take will be descriptive rather than prescriptive. The reasons for these changes are often difficult to pin down, but three forces for change I will explore towards the end of this paper are grammaticalisation, colloquialisation and Americanization. It is now possible to study and describe in some detail the recent changes in English grammatical usage that have taken place over a period of one generation: approximately 30 years. The changes that are observed over such a period are matters of increasing and decreasing frequency, but at the same time these changes are substantial. In this paper, I will report on some of the kinds of change that have been taking place in English grammar, focusing on the period 1961-1991. In the last section of this paper before the conclusion, I will briefly consider how far such gradual changes have implications for language teaching.

The Brown Family of Corpora How do we know about the changes that have been taking place? The answer lies in the existence of four carefully constructed electronic corpora: x The Brown Corpus: written American English text samples, published in 1961 x The Lancaster-Oslo/Bergen Corpus (‘LOB Corpus’): written British English 1961 x The Freiburg-Brown Corpus (‘Frown Corpus’): written American English 1991/2 x The Freiburg-Lancaster-Oslo/Bergen Corpus (‘FLOB Corpus’): written British English 1991 The names of these corpora (as well as their acronymic nicknames) derive from the universities at which they were compiled. For simplicity, we call these four the ‘Brown Family of Corpora’, and the relation between them is shown in Figure 3.1.

128

Chapter Eight

Figure 3.1 The Brown Family: the Brown, Frown, LOB and FLOB Corpora

The important point is that these corpora are all comparable corpora: that is, all four of them consist of a million words, and are equivalent collections of text samples: their sampling, taken from a wide range of text genres, is virtually equivalent. We might call them ‘clones’ of one another, except that they differ (a) in terms of date (1961 or 1991) and (b) in terms of their region of origin (American or British). Of course there are some limitations—notably that these corpora are restricted to written language—and some issues of detail mean that they are not 100% equivalent. Other varieties of world English—such as Australian English or Singapore English—are not accounted for. But for most purposes we can assume that the differences observed between these corpora reflect real changes in the use of the written language in the period 1961-1991.

Some Provisional Findings in Terms of Frequency In the following pages, I will explore some of the changes which the corpora reveal, focusing particularly on the English verb.

How English Grammar has been Changing

Modal would will can could may should must might shall ought (to) need + V TOTAL

Modal would will can could may should must might shall ought (to) need + V TOTAL

129

American English 1961 1992 3053 2868 2702 2402 2193 2160 1776 1655 1298 878 910 787 1018 668 665 635 267 150 69 49 40 35 13991 12287

Difference (%)* -6.1% -11.1% -1.5% -6.8% -32.4% -13.5% -34.4% -4.5% -43.8% -30.0% -12.5% -12.2%

Significance*

British English 1961 1991 3032 2682 2822 2708 2147 2213 1741 1767 1333 1100 1301 1148 1147 814 779 640 355 200 103 58 76 42.1 14836 13374

Difference (%)* -11.5% -4.0% +3.1% +1.5% -17.5% -11.8% -29.0% -17.8% -43.7% -43.7% -40.2% -9.9%

Significance*